Session

class sherpa.ui.utils.Session[source] [edit on github]

Bases: NoNewAttributesAfterInit

Methods Summary

add_model(modelclass[, args, kwargs])

Create a user-defined model class.

add_user_pars(modelname, parnames[, ...])

Add parameter information to a user model.

calc_chisqr([id])

Calculate the per-bin chi-squared statistic.

calc_stat([id])

Calculate the fit statistic for a data set.

calc_stat_info()

Display the statistic values for the current models.

clean()

Clear out the current Sherpa session.

conf(*args)

Estimate parameter confidence intervals using the confidence method.

confidence(*args)

Estimate parameter confidence intervals using the confidence method.

contour(*args, **kwargs)

Create a contour plot for an image data set.

contour_data([id, replot, overcontour])

Contour the values of an image data set.

contour_fit([id, replot, overcontour])

Contour the fit to a data set.

contour_fit_resid([id, replot, overcontour])

Contour the fit and the residuals to a data set.

contour_kernel([id, replot, overcontour])

Contour the kernel applied to the model of an image data set.

contour_model([id, replot, overcontour])

Create a contour plot of the model.

contour_psf([id, replot, overcontour])

Contour the PSF applied to the model of an image data set.

contour_ratio([id, replot, overcontour])

Contour the ratio of data to model.

contour_resid([id, replot, overcontour])

Contour the residuals of the fit.

contour_source([id, replot, overcontour])

Create a contour plot of the unconvolved spatial model.

copy_data(fromid, toid)

Copy a data set, creating a new identifier.

covar(*args)

Estimate parameter confidence intervals using the covariance method.

covariance(*args)

Estimate parameter confidence intervals using the covariance method.

create_model_component([typename, name])

Create a model component.

dataspace1d(start, stop[, step, numbins, ...])

Create the independent axis for a 1D data set.

dataspace2d(dims[, id, dstype])

Create the independent axis for a 2D data set.

delete_data([id])

Delete a data set by identifier.

delete_model([id])

Delete the model expression for a data set.

delete_model_component(name)

Delete a model component.

delete_psf([id])

Delete the PSF model for a data set.

fake([id, method])

Simulate a data set.

fit([id])

Fit a model to one or more data sets.

freeze(*args)

Fix model parameters so they are not changed by a fit.

get_cdf_plot()

Return the data used to plot the last CDF.

get_chisqr_plot([id, recalc])

Return the data used by plot_chisqr.

get_conf()

Return the confidence-interval estimation object.

get_conf_opt([name])

Return one or all of the options for the confidence interval method.

get_conf_results()

Return the results of the last conf run.

get_confidence_results()

Return the results of the last conf run.

get_covar()

Return the covariance estimation object.

get_covar_opt([name])

Return one or all of the options for the covariance method.

get_covar_results()

Return the results of the last covar run.

get_covariance_results()

Return the results of the last covar run.

get_data([id])

Return the data set by identifier.

get_data_contour([id, recalc])

Return the data used by contour_data.

get_data_contour_prefs()

Return the preferences for contour_data.

get_data_image([id])

Return the data used by image_data.

get_data_plot([id, recalc])

Return the data used by plot_data.

get_data_plot_prefs([id])

Return the preferences for plot_data.

get_default_id()

Return the default data set identifier.

get_delchi_plot([id, recalc])

Return the data used by plot_delchi.

get_dep([id, filter])

Return the dependent axis of a data set.

get_dims([id, filter])

Return the dimensions of the data set.

get_draws([id, otherids, niter, covar_matrix])

Run the pyBLoCXS MCMC algorithm.

get_error([id, filter])

Return the errors on the dependent axis of a data set.

get_filter([id])

Return the filter expression for a data set.

get_fit_contour([id, recalc])

Return the data used by contour_fit.

get_fit_plot([id, recalc])

Return the data used to create the fit plot.

get_fit_results()

Return the results of the last fit.

get_functions()

Return the functions provided by Sherpa.

get_indep([id])

Return the independent axes of a data set.

get_int_proj([par, id, otherids, recalc, ...])

Return the interval-projection object.

get_int_unc([par, id, otherids, recalc, ...])

Return the interval-uncertainty object.

get_iter_method_name()

Return the name of the iterative fitting scheme.

get_iter_method_opt([optname])

Return one or all options for the iterative-fitting scheme.

get_kernel_contour([id, recalc])

Return the data used by contour_kernel.

get_kernel_image([id])

Return the data used by image_kernel.

get_kernel_plot([id, recalc])

Return the data used by plot_kernel.

get_method([name])

Return an optimization method.

get_method_name()

Return the name of current Sherpa optimization method.

get_method_opt([optname])

Return one or all of the options for the current optimization method.

get_model([id])

Return the model expression for a data set.

get_model_autoassign_func()

Return the method used to create model component identifiers.

get_model_component(name)

Returns a model component given its name.

get_model_component_image(id[, model])

Return the data used by image_model_component.

get_model_component_plot(id[, model, recalc])

Return the data used to create the model-component plot.

get_model_contour([id, recalc])

Return the data used by contour_model.

get_model_contour_prefs()

Return the preferences for contour_model.

get_model_image([id])

Return the data used by image_model.

get_model_pars(model)

Return the names of the parameters of a model.

get_model_plot([id, recalc])

Return the data used to create the model plot.

get_model_plot_prefs([id])

Return the preferences for plot_model.

get_model_type(model)

Describe a model expression.

get_num_par([id])

Return the number of parameters in a model expression.

get_num_par_frozen([id])

Return the number of frozen parameters in a model expression.

get_num_par_thawed([id])

Return the number of thawed parameters in a model expression.

get_par(par)

Return a parameter of a model component.

get_pdf_plot()

Return the data used to plot the last PDF.

get_prior(par)

Return the prior function for a parameter (MCMC).

get_proj()

Return the confidence-interval estimation object.

get_proj_opt([name])

Return one or all of the options for the confidence interval method.

get_proj_results()

Return the results of the last proj run.

get_projection_results()

Return the results of the last proj run.

get_psf([id])

Return the PSF model defined for a data set.

get_psf_contour([id, recalc])

Return the data used by contour_psf.

get_psf_image([id])

Return the data used by image_psf.

get_psf_plot([id, recalc])

Return the data used by plot_psf.

get_pvalue_plot([null_model, alt_model, ...])

Return the data used by plot_pvalue.

get_pvalue_results()

Return the data calculated by the last plot_pvalue call.

get_ratio_contour([id, recalc])

Return the data used by contour_ratio.

get_ratio_image([id])

Return the data used by image_ratio.

get_ratio_plot([id, recalc])

Return the data used by plot_ratio.

get_reg_proj([par0, par1, id, otherids, ...])

Return the region-projection object.

get_reg_unc([par0, par1, id, otherids, ...])

Return the region-uncertainty object.

get_resid_contour([id, recalc])

Return the data used by contour_resid.

get_resid_image([id])

Return the data used by image_resid.

get_resid_plot([id, recalc])

Return the data used by plot_resid.

get_sampler()

Return the current MCMC sampler options.

get_sampler_name()

Return the name of the current MCMC sampler.

get_sampler_opt(opt)

Return an option of the current MCMC sampler.

get_scatter_plot()

Return the data used to plot the last scatter plot.

get_source([id])

Return the source model expression for a data set.

get_source_component_image(id[, model])

Return the data used by image_source_component.

get_source_component_plot(id[, model, recalc])

Return the data used by plot_source_component.

get_source_contour([id, recalc])

Return the data used by contour_source.

get_source_image([id])

Return the data used by image_source.

get_source_plot([id, recalc])

Return the data used to create the source plot.

get_split_plot()

Return the plot attributes for displays with multiple plots.

get_stat([name])

Return the fit statisic.

get_stat_info()

Return the statistic values for the current models.

get_stat_name()

Return the name of the current fit statistic.

get_staterror([id, filter])

Return the statistical error on the dependent axis of a data set.

get_syserror([id, filter])

Return the systematic error on the dependent axis of a data set.

get_trace_plot()

Return the data used to plot the last trace.

guess([id, model, limits, values])

Estimate the parameter values and ranges given the loaded data.

ignore([lo, hi])

Exclude data from the fit.

ignore_id(ids[, lo, hi])

Exclude data from the fit for a data set.

image_close()

Close the image viewer.

image_data([id, newframe, tile])

Display a data set in the image viewer.

image_deleteframes()

Delete all the frames open in the image viewer.

image_fit([id, newframe, tile, deleteframes])

Display the data, model, and residuals for a data set in the image viewer.

image_getregion([coord])

Return the region defined in the image viewer.

image_kernel([id, newframe, tile])

Display the 2D kernel for a data set in the image viewer.

image_model([id, newframe, tile])

Display the model for a data set in the image viewer.

image_model_component(id[, model, newframe, ...])

Display a component of the model in the image viewer.

image_open()

Start the image viewer.

image_psf([id, newframe, tile])

Display the 2D PSF model for a data set in the image viewer.

image_ratio([id, newframe, tile])

Display the ratio (data/model) for a data set in the image viewer.

image_resid([id, newframe, tile])

Display the residuals (data - model) for a data set in the image viewer.

image_setregion(reg[, coord])

Set the region to display in the image viewer.

image_source([id, newframe, tile])

Display the source expression for a data set in the image viewer.

image_source_component(id[, model, ...])

Display a component of the source expression in the image viewer.

image_xpaget(arg)

Return the result of an XPA call to the image viewer.

image_xpaset(arg[, data])

Return the result of an XPA call to the image viewer.

int_proj(par[, id, otherids, replot, fast, ...])

Calculate and plot the fit statistic versus fit parameter value.

int_unc(par[, id, otherids, replot, min, ...])

Calculate and plot the fit statistic versus fit parameter value.

link(par, val)

Link a parameter to a value.

list_data_ids()

List the identifiers for the loaded data sets.

list_functions([outfile, clobber])

Display the functions provided by Sherpa.

list_iter_methods()

List the iterative fitting schemes.

list_methods()

List the optimization methods.

list_model_components()

List the names of all the model components.

list_model_ids()

List of all the data sets with a source expression.

list_models([show])

List the available model types.

list_priors()

Return the priors set for model parameters, if any.

list_psf_ids()

List of all the data sets with a PSF.

list_samplers()

List the MCMC samplers.

list_stats()

List the fit statistics.

load_arrays(id, *args)

Create a data set from array values.

load_conv(modelname, filename_or_model, ...)

Load a 1D convolution model.

load_data(id[, filename, ncols, colkeys, ...])

Load a data set from an ASCII file.

load_filter(id[, filename, ignore, ncols])

Load the filter array from an ASCII file and add to a data set.

load_psf(modelname, filename_or_model, ...)

Create a PSF model.

load_staterror(id[, filename, ncols])

Load the statistical errors from an ASCII file.

load_syserror(id[, filename, ncols])

Load the systematic errors from an ASCII file.

load_table_model(modelname, filename[, ...])

Load ASCII tabular data and use it as a model component.

load_template_interpolator(name, ...)

Set the template interpolation scheme.

load_template_model(modelname, templatefile)

Load a set of templates and use it as a model component.

load_user_model(func, modelname[, filename, ...])

Create a user-defined model.

load_user_stat(statname, calc_stat_func[, ...])

Create a user-defined statistic.

normal_sample([num, sigma, correlate, id, ...])

Sample the fit statistic by taking the parameter values from a normal distribution.

notice([lo, hi])

Include data in the fit.

notice_id(ids[, lo, hi])

Include data from the fit for a data set.

paramprompt([val])

Should the user be asked for the parameter values when creating a model?

plot(*args, **kwargs)

Create one or more plot types.

plot_cdf(points[, name, xlabel, replot, ...])

Plot the cumulative density function of an array of values.

plot_chisqr([id, replot, overplot, clearwindow])

Plot the chi-squared value for each point in a data set.

plot_data([id, replot, overplot, clearwindow])

Plot the data values.

plot_delchi([id, replot, overplot, clearwindow])

Plot the ratio of residuals to error for a data set.

plot_fit([id, replot, overplot, clearwindow])

Plot the fit results (data, model) for a data set.

plot_fit_delchi([id, replot, overplot, ...])

Plot the fit results, and the residuals, for a data set.

plot_fit_ratio([id, replot, overplot, ...])

Plot the fit results, and the ratio of data to model, for a data set.

plot_fit_resid([id, replot, overplot, ...])

Plot the fit results, and the residuals, for a data set.

plot_kernel([id, replot, overplot, clearwindow])

Plot the 1D kernel applied to a data set.

plot_model([id, replot, overplot, clearwindow])

Plot the model for a data set.

plot_model_component(id[, model, replot, ...])

Plot a component of the model for a data set.

plot_pdf(points[, name, xlabel, bins, ...])

Plot the probability density function of an array of values.

plot_psf([id, replot, overplot, clearwindow])

Plot the 1D PSF model applied to a data set.

plot_pvalue(null_model, alt_model[, ...])

Compute and plot a histogram of likelihood ratios by simulating data.

plot_ratio([id, replot, overplot, clearwindow])

Plot the ratio of data to model for a data set.

plot_resid([id, replot, overplot, clearwindow])

Plot the residuals (data - model) for a data set.

plot_scatter(x, y[, name, xlabel, ylabel, ...])

Create a scatter plot.

plot_source([id, replot, overplot, clearwindow])

Plot the source expression for a data set.

plot_source_component(id[, model, replot, ...])

Plot a component of the source expression for a data set.

plot_trace(points[, name, xlabel, replot, ...])

Create a trace plot of row number versus value.

proj(*args)

Estimate parameter confidence intervals using the projection method.

projection(*args)

Estimate parameter confidence intervals using the projection method.

reg_proj(par0, par1[, id, otherids, replot, ...])

Plot the statistic value as two parameters are varied.

reg_unc(par0, par1[, id, otherids, replot, ...])

Plot the statistic value as two parameters are varied.

reset([model, id])

Reset the model parameters to their default settings.

restore([filename])

Load in a Sherpa session from a file.

save([filename, clobber])

Save the current Sherpa session to a file.

save_arrays(filename, args[, fields, ...])

Write a list of arrays to an ASCII file.

save_data(id[, filename, fields, sep, ...])

Save the data to a file.

save_delchi(id[, filename, clobber, sep, ...])

Save the ratio of residuals (data-model) to error to a file.

save_error(id[, filename, clobber, sep, ...])

Save the errors to a file.

save_filter(id[, filename, clobber, sep, ...])

Save the filter array to a file.

save_model(id[, filename, clobber, sep, ...])

Save the model values to a file.

save_resid(id[, filename, clobber, sep, ...])

Save the residuals (data-model) to a file.

save_source(id[, filename, clobber, sep, ...])

Save the model values to a file.

save_staterror(id[, filename, clobber, sep, ...])

Save the statistical errors to a file.

save_syserror(id[, filename, clobber, sep, ...])

Save the statistical errors to a file.

set_conf_opt(name, val)

Set an option for the confidence interval method.

set_covar_opt(name, val)

Set an option for the covariance method.

set_data(id[, data])

Set a data set.

set_default_id(id)

Set the default data set identifier.

set_dep(id[, val])

Set the dependent axis of a data set.

set_filter(id[, val, ignore])

Set the filter array of a data set.

set_full_model(id[, model])

Define the convolved model expression for a data set.

set_iter_method(meth)

Set the iterative-fitting scheme used in the fit.

set_iter_method_opt(optname, val)

Set an option for the iterative-fitting scheme.

set_method(meth)

Set the optimization method.

set_method_opt(optname, val)

Set an option for the current optimization method.

set_model(id[, model])

Set the source model expression for a data set.

set_model_autoassign_func([func])

Set the method used to create model component identifiers.

set_par(par[, val, min, max, frozen])

Set the value, limits, or behavior of a model parameter.

set_prior(par, prior)

Set the prior function to use with a parameter.

set_proj_opt(name, val)

Set an option for the projection method.

set_psf(id[, psf])

Add a PSF model to a data set.

set_sampler(sampler)

Set the MCMC sampler.

set_sampler_opt(opt, value)

Set an option for the current MCMC sampler.

set_source(id[, model])

Set the source model expression for a data set.

set_stat(stat)

Set the statistical method.

set_staterror(id[, val, fractional])

Set the statistical errors on the dependent axis of a data set.

set_syserror(id[, val, fractional])

Set the systematic errors on the dependent axis of a data set.

set_xlinear([plottype])

New plots will display a linear X axis.

set_xlog([plottype])

New plots will display a logarithmically-scaled X axis.

set_ylinear([plottype])

New plots will display a linear Y axis.

set_ylog([plottype])

New plots will display a logarithmically-scaled Y axis.

show_all([id, outfile, clobber])

Report the current state of the Sherpa session.

show_conf([outfile, clobber])

Display the results of the last conf evaluation.

show_covar([outfile, clobber])

Display the results of the last covar evaluation.

show_data([id, outfile, clobber])

Summarize the available data sets.

show_filter([id, outfile, clobber])

Show any filters applied to a data set.

show_fit([outfile, clobber])

Summarize the fit results.

show_kernel([id, outfile, clobber])

Display any kernel applied to a data set.

show_method([outfile, clobber])

Display the current optimization method and options.

show_model([id, outfile, clobber])

Display the model expression used to fit a data set.

show_proj([outfile, clobber])

Display the results of the last proj evaluation.

show_psf([id, outfile, clobber])

Display any PSF model applied to a data set.

show_source([id, outfile, clobber])

Display the source model expression for a data set.

show_stat([outfile, clobber])

Display the current fit statistic.

simulfit([id])

Fit a model to one or more data sets.

t_sample([num, dof, id, otherids, numcores])

Sample the fit statistic by taking the parameter values from a Student's t-distribution.

thaw(*args)

Allow model parameters to be varied during a fit.

uniform_sample([num, factor, id, otherids, ...])

Sample the fit statistic by taking the parameter values from an uniform distribution.

unlink(par)

Unlink a parameter value.

unpack_arrays(*args)

Create a sherpa data object from arrays of data.

unpack_data(filename[, ncols, colkeys, ...])

Create a sherpa data object from an ASCII file.

Methods Documentation

add_model(modelclass, args=(), kwargs={})[source] [edit on github]

Create a user-defined model class.

Create a model from a class. The name of the class can then be used to create model components - e.g. with set_model or create_model_component - as with any existing Sherpa model.

Parameters
  • modelclass – A class derived from sherpa.models.model.ArithmeticModel. This class defines the functional form and the parameters of the model.

  • args – Arguments for the class constructor.

  • kwargs – Keyword arguments for the class constructor.

See also

create_model_component

Create a model component.

list_models

List the available model types.

load_table_model

Load tabular data and use it as a model component.

load_user_model

Create a user-defined model.

set_model

Set the source model expression for a data set.

Notes

The load_user_model function is designed to make it easy to add a model, but the interface is not the same as the existing models (such as having to call both load_user_model and add_user_pars for each new instance). The add_model function is used to add a model as a Python class, which is more work to set up, but then acts the same way as the existing models.

Examples

The following example creates a model type called “mygauss1d” which will behave excatly the same as the existing “gauss1d” model. Normally the class used with add_model would add new functionality.

>>> from sherpa.models import Gauss1D
>>> class MyGauss1D(Gauss1D):
...     pass
...
>>> add_model(MyGauss1D)
>>> set_source(mygauss1d.g1 + mygauss1d.g2)
add_user_pars(modelname, parnames, parvals=None, parmins=None, parmaxs=None, parunits=None, parfrozen=None)[source] [edit on github]

Add parameter information to a user model.

Parameters
  • modelname (str) – The name of the user model (created by load_user_model).

  • parnames (array of str) – The names of the parameters. The order of all the parameter arrays must match that expected by the model function (the first argument to load_user_model).

  • parvals (array of number, optional) – The default values of the parameters. If not given each parameter is set to 0.

  • parmins (array of number, optional) – The minimum values of the parameters (hard limit). The default value is -3.40282e+38.

  • parmaxs (array of number, optional) – The maximum values of the parameters (hard limit). The default value is 3.40282e+38.

  • parunits (array of str, optional) – The units of the parameters. This is only used in screen output (i.e. is informational in nature).

  • parfrozen (array of bool, optional) – Should each parameter be frozen. The default is that all parameters are thawed.

See also

add_model

Create a user-defined model class.

load_user_model

Create a user-defined model.

set_par

Set the value, limits, or behavior of a model parameter.

Notes

The parameters must be specified in the order that the function expects. That is, if the function has two parameters, pars[0]=’slope’ and pars[1]=’y_intercept’, then the call to add_user_pars must use the order [“slope”, “y_intercept”].

Examples

Create a user model for the function profile called “myprof”, which has two parameters called “core” and “ampl”, both of which will start with a value of 0.

>>> load_user_model(profile, "myprof")
>>> add_user_pars("myprof", ["core", "ampl"])

Set the starting values, minimum values, and whether or not the parameter is frozen by default, for the “prof” model:

>>> pnames = ["core", "ampl", "intflag"]
>>> pvals = [10, 200, 1]
>>> pmins = [0.01, 0, 0]
>>> pfreeze = [False, False, True]
>>> add_user_pars("prof", pnames, pvals,
...               parmins=pmins, parfrozen=pfreeze)
calc_chisqr(id=None, *otherids)[source] [edit on github]

Calculate the per-bin chi-squared statistic.

Evaluate the model for one or more data sets, compare it to the data using the current statistic, and return an array of chi-squared values for each bin. No fitting is done, as the current model parameter, and any filters, are used.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • *otherids (int or str, optional) – Other data sets to use in the calculation.

Returns

chisq – The chi-square value for each bin of the data, using the current statistic (as set by set_stat). A value of None is returned if the statistic is not a chi-square distribution.

Return type

array or None

See also

calc_stat

Calculate the fit statistic for a data set.

calc_stat_info

Display the statistic values for the current models.

set_stat

Set the statistical method.

Notes

The output array length equals the sum of the arrays lengths of the requested data sets.

Examples

When called with no arguments, the return value is the chi-squared statistic for each bin in the data sets which have a defined model.

>>> calc_chisqr()

Supplying a specific data set ID to calc_chisqr - such as “1” or “src” - will return the chi-squared statistic array for only that data set.

>>> calc_chisqr(1)
>>> calc_chisqr("src")

Restrict the calculation to just datasets 1 and 3:

>>> calc_chisqr(1, 3)
calc_stat(id=None, *otherids)[source] [edit on github]

Calculate the fit statistic for a data set.

Evaluate the model for one or more data sets, compare it to the data using the current statistic, and return the value. No fitting is done, as the current model parameter, and any filters, are used.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • *otherids (int or str, optional) – Other data sets to use in the calculation.

Returns

stat – The current statistic value.

Return type

number

See also

calc_chisqr

Calculate the per-bin chi-squared statistic.

calc_stat_info

Display the statistic values for the current models.

set_stat

Set the statistical method.

Examples

Calculate the statistic for the model and data in the default data set:

>>> stat = calc_stat()

Find the statistic for data set 3:

>>> stat = calc_stat(3)

When fitting to multiple data sets, you can get the contribution to the total fit statistic from only one data set, or from several by listing the datasets explicitly. The following finds the contribution from the data sets labelled “core” and “jet”:

>>> stat = calc_stat("core", "jet")

Calculate the statistic value using two different statistics:

>>> set_stat('cash')
>>> s1 = calc_stat()
>>> set_stat('cstat')
>>> s2 = calc_stat()
calc_stat_info()[source] [edit on github]

Display the statistic values for the current models.

Displays the statistic value for each data set, and the combined fit, using the current set of models, parameters, and ranges. The output is printed to stdout, and so is intended for use in interactive analysis. The get_stat_info function returns the same information but as an array of Python structures.

See also

calc_stat

Calculate the fit statistic for a data set.

get_stat_info

Return the statistic values for the current models.

Notes

If a fit to a particular data set has not been made, or values - such as parameter settings, the noticed data range, or choice of statistic - have been changed since the last fit, then the results for that data set may not be meaningful and will therefore bias the results for the simultaneous results.

The information returned by calc_stat_info includes:

Dataset

The dataset identifier (or identifiers).

Statistic

The name of the statistic used to calculate the results.

Fit statistic value

The current fit statistic value.

Data points

The number of bins used in the fit.

Degrees of freedom

The number of bins minus the number of thawed parameters.

Some fields are only returned for a subset of statistics:

Probability (Q-value)

A measure of the probability that one would observe the reduced statistic value, or a larger value, if the assumed model is true and the best-fit model parameters are the true parameter values.

Reduced statistic

The fit statistic value divided by the number of degrees of freedom.

Examples

>>> calc_stat_info()
clean()[source] [edit on github]

Clear out the current Sherpa session.

The clean function removes all data sets and model assignments, and restores the default settings for the optimisation and fit statistic.

See also

save

Save the current Sherpa session to a file.

restore

Load in a Sherpa session from a file.

sherpa.astro.ui.save_all

Save the Sherpa session as an ASCII file.

Examples

>>> clean()
conf(*args)[source] [edit on github]

Estimate parameter confidence intervals using the confidence method.

The conf command computes confidence interval bounds for the specified model parameters in the dataset. A given parameter’s value is varied along a grid of values while the values of all the other thawed parameters are allowed to float to new best-fit values. The get_conf and set_conf_opt commands can be used to configure the error analysis; an example being changing the ‘sigma’ field to 1.6 (i.e. 90%) from its default value of 1. The output from the routine is displayed on screen, and the get_conf_results routine can be used to retrieve the results.

Parameters
  • id (int or str, optional) – The data set, or sets, that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • parameter (sherpa.models.parameter.Parameter, optional) – The default is to calculate the confidence limits on all thawed parameters of the model, or models, for all the data sets. The evaluation can be restricted by listing the parameters to use. Note that each parameter should be given as a separate argument, rather than as a list. For example conf(g1.ampl, g1.sigma).

  • model (sherpa.models.model.Model, optional) – Select all the thawed parameters in the model.

See also

covar

Estimate the confidence intervals using the covariance method.

get_conf

Return the confidence-interval estimation object.

get_conf_results

Return the results of the last conf run.

int_proj

Plot the statistic value as a single parameter is varied.

int_unc

Plot the statistic value as a single parameter is varied.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

set_conf_opt

Set an option of the conf estimation object.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with multiple ids or parameters values, the order is unimportant, since any argument that is not defined as a model parameter is assumed to be a data id.

The conf function is different to covar, in that in that all other thawed parameters are allowed to float to new best-fit values, instead of being fixed to the initial best-fit values as they are in covar. While conf is more general (e.g. allowing the user to examine the parameter space away from the best-fit point), it is in the strictest sense no more accurate than covar for determining confidence intervals.

The conf function is a replacement for the proj function, which uses a different algorithm to estimate parameter confidence limits.

An estimated confidence interval is accurate if and only if:

  1. the chi^2 or logL surface in parameter space is approximately shaped like a multi-dimensional paraboloid, and

  2. the best-fit point is sufficiently far from parameter space boundaries.

One may determine if these conditions hold, for example, by plotting the fit statistic as a function of each parameter’s values (the curve should approximate a parabola) and by examining contour plots of the fit statistics made by varying the values of two parameters at a time (the contours should be elliptical, and parameter space boundaries should be no closer than approximately 3 sigma from the best-fit point). The int_proj and reg_proj commands may be used for this.

If either of the conditions given above does not hold, then the output from conf may be meaningless except to give an idea of the scale of the confidence intervals. To accurately determine the confidence intervals, one would have to reparameterize the model, use Monte Carlo simulations, or Bayesian methods.

As the calculation can be computer intensive, the default behavior is to use all available CPU cores to speed up the analysis. This can be changed be varying the numcores option - or setting parallel to False - either with set_conf_opt or get_conf.

As conf estimates intervals for each parameter independently, the relationship between sigma and the change in statistic value delta_S can be particularly simple: sigma = the square root of delta_S for statistics sampled from the chi-square distribution and for the Cash statistic, and is approximately equal to the square root of (2 * delta_S) for fits based on the general log-likelihood. The default setting is to calculate the one-sigma interval, which can be changed with the sigma option to set_conf_opt or get_conf.

The limit calculated by conf is basically a 1-dimensional root in the translated coordinate system (translated by the value of the statistic at the minimum plus sigma^2). The Taylor series expansion of the multi-dimensional function at the minimum is:

f(x + dx) ~ f(x) + grad( f(x) )^T dx + dx^T Hessian( f(x) ) dx + ...

where x is understood to be the n-dimensional vector representing the free parameters to be fitted and the super-script ‘T’ is the transpose of the row-vector. At or near the minimum, the gradient of the function is zero or negligible, respectively. So the leading term of the expansion is quadratic. The best root finding algorithm for a curve which is approximately parabolic is Muller’s method [1]_. Muller’s method is a generalization of the secant method [2]_: the secant method is an iterative root finding method that approximates the function by a straight line through two points, whereas Muller’s method is an iterative root finding method that approxmiates the function by a quadratic polynomial through three points.

Three data points are the minimum input to Muller’s root finding method. The first point to be submitted to the Muller’s root finding method is the point at the minimum. To strategically choose the other two data points, the confidence function uses the output from covariance as the second data point. To generate the third data points for the input to Muller’s root finding method, the secant root finding method is used since it only requires two data points to generate the next best approximation of the root.

However, there are cases where conf cannot locate the root even though the root is bracketed within an interval (perhaps due to the bad resolution of the data). In such cases, when the option openinterval is set to False (which is the default), the routine will print a warning message about not able to find the root within the set tolerance and the function will return the average of the open interval which brackets the root. If openinterval is set to True then conf will print the minimal open interval which brackets the root (not to be confused with the lower and upper bound of the confidence interval). The most accurate thing to do is to return an open interval where the root is localized/bracketed rather then the average of the open interval (since the average of the interval is not a root within the specified tolerance).

References

1

Muller, David E., “A Method for Solving Algebraic Equations Using an Automatic Computer,” MTAC, 10 (1956), 208-215.

2

Numerical Recipes in Fortran, 2nd edition, 1986, Press et al., p. 347

Examples

Evaluate confidence intervals for all thawed parameters in all data sets with an associated source model. The results are then stored in the variable res.

>>> conf()
>>> res = get_conf_results()

Only evaluate the parameters associated with data set 2:

>>> conf(2)

Only evaluate the intervals for the pos.xpos and pos.ypos parameters:

>>> conf(pos.xpos, pos.ypos)

Change the limits to be 1.6 sigma (90%) rather than the default 1 sigma.

>>> set_conf_opt('sigma', 1.6)
>>> conf()

Only evaluate the clus.kt parameter for the data sets with identifiers “obs1”, “obs5”, and “obs6”. This will still use the 1.6 sigma setting from the previous run.

>>> conf("obs1", "obs5", "obs6", clus.kt)

Only use two cores when evaluating the errors for the parameters used in the model for data set 3:

>>> set_conf_opt('numcores', 2)
>>> conf(3)

Estimate the errors for all the thawed parameters from the line model and the clus.kt parameter for datasets 1, 3, and 4:

>>> conf(1, 3, 4, line, clus.kt)
confidence(*args) [edit on github]

Estimate parameter confidence intervals using the confidence method.

The conf command computes confidence interval bounds for the specified model parameters in the dataset. A given parameter’s value is varied along a grid of values while the values of all the other thawed parameters are allowed to float to new best-fit values. The get_conf and set_conf_opt commands can be used to configure the error analysis; an example being changing the ‘sigma’ field to 1.6 (i.e. 90%) from its default value of 1. The output from the routine is displayed on screen, and the get_conf_results routine can be used to retrieve the results.

Parameters
  • id (int or str, optional) – The data set, or sets, that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • parameter (sherpa.models.parameter.Parameter, optional) – The default is to calculate the confidence limits on all thawed parameters of the model, or models, for all the data sets. The evaluation can be restricted by listing the parameters to use. Note that each parameter should be given as a separate argument, rather than as a list. For example conf(g1.ampl, g1.sigma).

  • model (sherpa.models.model.Model, optional) – Select all the thawed parameters in the model.

See also

covar

Estimate the confidence intervals using the covariance method.

get_conf

Return the confidence-interval estimation object.

get_conf_results

Return the results of the last conf run.

int_proj

Plot the statistic value as a single parameter is varied.

int_unc

Plot the statistic value as a single parameter is varied.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

set_conf_opt

Set an option of the conf estimation object.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with multiple ids or parameters values, the order is unimportant, since any argument that is not defined as a model parameter is assumed to be a data id.

The conf function is different to covar, in that in that all other thawed parameters are allowed to float to new best-fit values, instead of being fixed to the initial best-fit values as they are in covar. While conf is more general (e.g. allowing the user to examine the parameter space away from the best-fit point), it is in the strictest sense no more accurate than covar for determining confidence intervals.

The conf function is a replacement for the proj function, which uses a different algorithm to estimate parameter confidence limits.

An estimated confidence interval is accurate if and only if:

  1. the chi^2 or logL surface in parameter space is approximately shaped like a multi-dimensional paraboloid, and

  2. the best-fit point is sufficiently far from parameter space boundaries.

One may determine if these conditions hold, for example, by plotting the fit statistic as a function of each parameter’s values (the curve should approximate a parabola) and by examining contour plots of the fit statistics made by varying the values of two parameters at a time (the contours should be elliptical, and parameter space boundaries should be no closer than approximately 3 sigma from the best-fit point). The int_proj and reg_proj commands may be used for this.

If either of the conditions given above does not hold, then the output from conf may be meaningless except to give an idea of the scale of the confidence intervals. To accurately determine the confidence intervals, one would have to reparameterize the model, use Monte Carlo simulations, or Bayesian methods.

As the calculation can be computer intensive, the default behavior is to use all available CPU cores to speed up the analysis. This can be changed be varying the numcores option - or setting parallel to False - either with set_conf_opt or get_conf.

As conf estimates intervals for each parameter independently, the relationship between sigma and the change in statistic value delta_S can be particularly simple: sigma = the square root of delta_S for statistics sampled from the chi-square distribution and for the Cash statistic, and is approximately equal to the square root of (2 * delta_S) for fits based on the general log-likelihood. The default setting is to calculate the one-sigma interval, which can be changed with the sigma option to set_conf_opt or get_conf.

The limit calculated by conf is basically a 1-dimensional root in the translated coordinate system (translated by the value of the statistic at the minimum plus sigma^2). The Taylor series expansion of the multi-dimensional function at the minimum is:

f(x + dx) ~ f(x) + grad( f(x) )^T dx + dx^T Hessian( f(x) ) dx + ...

where x is understood to be the n-dimensional vector representing the free parameters to be fitted and the super-script ‘T’ is the transpose of the row-vector. At or near the minimum, the gradient of the function is zero or negligible, respectively. So the leading term of the expansion is quadratic. The best root finding algorithm for a curve which is approximately parabolic is Muller’s method [1]_. Muller’s method is a generalization of the secant method [2]_: the secant method is an iterative root finding method that approximates the function by a straight line through two points, whereas Muller’s method is an iterative root finding method that approxmiates the function by a quadratic polynomial through three points.

Three data points are the minimum input to Muller’s root finding method. The first point to be submitted to the Muller’s root finding method is the point at the minimum. To strategically choose the other two data points, the confidence function uses the output from covariance as the second data point. To generate the third data points for the input to Muller’s root finding method, the secant root finding method is used since it only requires two data points to generate the next best approximation of the root.

However, there are cases where conf cannot locate the root even though the root is bracketed within an interval (perhaps due to the bad resolution of the data). In such cases, when the option openinterval is set to False (which is the default), the routine will print a warning message about not able to find the root within the set tolerance and the function will return the average of the open interval which brackets the root. If openinterval is set to True then conf will print the minimal open interval which brackets the root (not to be confused with the lower and upper bound of the confidence interval). The most accurate thing to do is to return an open interval where the root is localized/bracketed rather then the average of the open interval (since the average of the interval is not a root within the specified tolerance).

References

1

Muller, David E., “A Method for Solving Algebraic Equations Using an Automatic Computer,” MTAC, 10 (1956), 208-215.

2

Numerical Recipes in Fortran, 2nd edition, 1986, Press et al., p. 347

Examples

Evaluate confidence intervals for all thawed parameters in all data sets with an associated source model. The results are then stored in the variable res.

>>> conf()
>>> res = get_conf_results()

Only evaluate the parameters associated with data set 2:

>>> conf(2)

Only evaluate the intervals for the pos.xpos and pos.ypos parameters:

>>> conf(pos.xpos, pos.ypos)

Change the limits to be 1.6 sigma (90%) rather than the default 1 sigma.

>>> set_conf_opt('sigma', 1.6)
>>> conf()

Only evaluate the clus.kt parameter for the data sets with identifiers “obs1”, “obs5”, and “obs6”. This will still use the 1.6 sigma setting from the previous run.

>>> conf("obs1", "obs5", "obs6", clus.kt)

Only use two cores when evaluating the errors for the parameters used in the model for data set 3:

>>> set_conf_opt('numcores', 2)
>>> conf(3)

Estimate the errors for all the thawed parameters from the line model and the clus.kt parameter for datasets 1, 3, and 4:

>>> conf(1, 3, 4, line, clus.kt)
contour(*args, **kwargs)[source] [edit on github]

Create a contour plot for an image data set.

Create one or more contour plots, depending on the arguments it is set: a plot type, followed by an optional data set identifier, and this can be repeated. If no data set identifier is given for a plot type, the default identifier - as returned by get_default_id - is used. This is for 2D data sets.

Changed in version 4.12.2: Keyword arguments, such as alpha, can be sent to each plot.

Raises

sherpa.utils.err.DataErr – The data set does not support the requested plot type.

See also

contour_data

Contour the values of an image data set.

contour_fit

Contour the fit to a data set.

contour_fit_resid

Contour the fit and the residuals to a data set.

contour_kernel

Contour the kernel applied to the model of an image data set.

contour_model

Contour the values of the model, including any PSF.

contour_psf

Contour the PSF applied to the model of an image data set.

contour_ratio

Contour the ratio of data to model.

contour_resid

Contour the residuals of the fit.

contour_source

Contour the values of the model, without any PSF.

get_default_id

Return the default data set identifier.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Notes

The supported plot types depend on the data set type, and include the following list. There are also individual functions, with contour_ prepended to the plot type, such as contour_data and the contour_fit_resid variant:

data

The data.

fit

Contours of the data and the source model.

fit_resid

Two plots: the first is the contours of the data and the source model and the second is the residuals.

kernel

The kernel.

model

The source model including any PSF convolution set by set_psf.

psf

The PSF.

ratio

Contours of the ratio image, formed by dividing the data by the model.

resid

Contours of the residual image, formed by subtracting the model from the data.

source

The source model (without any PSF convolution set by set_psf).

The keyword arguments are sent to each plot (so care must be taken to ensure they are valid for all plots).

Examples

>>> contour('data')
>>> contour('data', 1, 'data', 2)
>>> contour('data', 'model')
>>> contour('data', 'model', 'fit', 'resid')
>>> contour('data', 'model', alpha=0.7)
contour_data(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the values of an image data set.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_data. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_data_contour

Return the data used by contour_data.

get_data_contour_prefs

Return the preferences for contour_data.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Examples

Plot the data from the default data set:

>>> contour_data()

Contour the data and then overplot the data from the second data set:

>>> contour_data()
>>> contour_data(2, overcontour=True)
contour_fit(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the fit to a data set.

Overplot the model - including any PSF - on the data. The preferences are the same as contour_data and contour_model.

Parameters
  • id (int or str, optional) – The data set that provides the data and model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_fit. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_fit_contour

Return the data used by contour_fit.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Examples

Plot the fit for the default data set:

>>> contour_fit()

Overplot the fit to data set ‘s2’ on that of the default data set:

>>> contour_fit()
>>> contour_fit('s2', overcontour=True)
contour_fit_resid(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the fit and the residuals to a data set.

Overplot the model - including any PSF - on the data. In a separate plot contour the residuals. The preferences are the same as contour_data and contour_model.

Parameters
  • id (int or str, optional) – The data set that provides the data and model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_fit_resid. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_fit_contour

Return the data used by contour_fit.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

contour_fit

Contour the fit to a data set.

contour_resid

Contour the residuals of the fit.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Examples

Plot the fit and residuals for the default data set:

>>> contour_fit_resid()
contour_kernel(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the kernel applied to the model of an image data set.

If the data set has no PSF applied to it, the model is displayed.

Parameters
  • id (int or str, optional) – The data set that provides the model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_kernel. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_psf_contour

Return the data used by contour_psf.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

contour_psf

Contour the PSF applied to the model of an image data set.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

set_psf

Add a PSF model to a data set.

contour_model(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Create a contour plot of the model.

Displays a contour plot of the values of the model, evaluated on the data, including any PSF kernel convolution (if set).

Parameters
  • id (int or str, optional) – The data set that provides the model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_model. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_model_contour

Return the data used by contour_model.

get_model_contour_prefs

Return the preferences for contour_model.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

contour_source

Create a contour plot of the unconvolved spatial model.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

set_psf

Add a PSF model to a data set.

Examples

Plot the model from the default data set:

>>> contour_model()

Compare the model without and with the PSF component, for the “img” data set:

>>> contour_source("img")
>>> contour_model("img", overcontour=True)
contour_psf(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the PSF applied to the model of an image data set.

If the data set has no PSF applied to it, the model is displayed.

Parameters
  • id (int or str, optional) – The data set that provides the model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_psf. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_psf_contour

Return the data used by contour_psf.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

contour_kernel

Contour the kernel applied to the model of an image data set.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

set_psf

Add a PSF model to a data set.

contour_ratio(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the ratio of data to model.

The ratio image is formed by dividing the data by the current model, including any PSF. The preferences are the same as contour_data.

Parameters
  • id (int or str, optional) – The data set that provides the data and model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_ratio. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_ratio_contour

Return the data used by contour_ratio.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Examples

Plot the ratio from the default data set:

>>> contour_ratio()

Overplot the ratio on the residuals:

>>> contour_resid('img')
>>> contour_ratio('img', overcontour=True)
contour_resid(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Contour the residuals of the fit.

The residuals are formed by subtracting the current model - including any PSF - from the data. The preferences are the same as contour_data.

Parameters
  • id (int or str, optional) – The data set that provides the data and model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_resid. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_resid_contour

Return the data used by contour_resid.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

Examples

Plot the residuals from the default data set:

>>> contour_resid()

Overplot the residuals on the model:

>>> contour_model('img')
>>> contour_resid('img', overcontour=True)
contour_source(id=None, replot=False, overcontour=False, **kwargs)[source] [edit on github]

Create a contour plot of the unconvolved spatial model.

Displays a contour plot of the values of the model, evaluated on the data, without any PSF kernel convolution applied. The preferences are the same as contour_model.

Parameters
  • id (int or str, optional) – The data set that provides the model. If not given then the default identifier is used, as returned by get_default_id.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to contour_source. The default is False.

  • overcontour (bool, optional) – If True then add the data to an existing plot, otherwise create a new contour plot. The default is False.

See also

get_source_contour

Return the data used by contour_source.

get_default_id

Return the default data set identifier.

contour

Create one or more plot types.

contour_model

Create a contour plot of the model.

sherpa.astro.ui.set_coord

Set the coordinate system to use for image analysis.

set_psf

Add a PSF model to a data set.

Examples

Plot the model from the default data set:

>>> contour_source()

Compare the model without and with the PSF component, for the “img” data set:

>>> contour_model("img")
>>> contour_source("img", overcontour=True)
copy_data(fromid, toid)[source] [edit on github]

Copy a data set, creating a new identifier.

After copying the data set, any changes made to the original data set (that is, the fromid identifier) will not be reflected in the new (the toid identifier) data set.

Parameters
  • fromid (int or str) – The input data set.

  • toid (int or str) – The output data set.

Raises

sherpa.utils.err.IdentifierErr – If there is no data set with a fromid identifier.

Examples

>>> copy_data(1, 2)

Rename the data set with identifier 2 to “orig”, and then delete the old data set:

>>> copy_data(2, "orig")
>>> delete_data(2)
covar(*args)[source] [edit on github]

Estimate parameter confidence intervals using the covariance method.

The covar command computes confidence interval bounds for the specified model parameters in the dataset, using the covariance matrix of the statistic. The get_covar and set_covar_opt commands can be used to configure the error analysis; an example being changing the sigma field to 1.6 (i.e. 90%) from its default value of 1. The output from the routine is displayed on screen, and the get_covar_results routine can be used to retrieve the results.

Parameters
  • id (int or str, optional) – The data set, or sets, that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • parameter (sherpa.models.parameter.Parameter, optional) – The default is to calculate the confidence limits on all thawed parameters of the model, or models, for all the data sets. The evaluation can be restricted by listing the parameters to use. Note that each parameter should be given as a separate argument, rather than as a list. For example covar(g1.ampl, g1.sigma).

  • model (sherpa.models.model.Model, optional) – Select all the thawed parameters in the model.

See also

covar

Estimate the confidence intervals using the confidence method.

get_covar

Return the covariance estimation object.

get_covar_results

Return the results of the last covar run.

int_proj

Plot the statistic value as a single parameter is varied.

int_unc

Plot the statistic value as a single parameter is varied.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

set_covar_opt

Set an option of the covar estimation object.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with multiple ids or parameters values, the order is unimportant, since any argument that is not defined as a model parameter is assumed to be a data id.

The covar command is different to conf, in that in that all other thawed parameters are fixed, rather than being allowed to float to new best-fit values. While conf is more general (e.g. allowing the user to examine the parameter space away from the best-fit point), it is in the strictest sense no more accurate than covar for determining confidence intervals.

An estimated confidence interval is accurate if and only if:

  1. the chi^2 or logL surface in parameter space is approximately shaped like a multi-dimensional paraboloid, and

  2. the best-fit point is sufficiently far from parameter space boundaries.

One may determine if these conditions hold, for example, by plotting the fit statistic as a function of each parameter’s values (the curve should approximate a parabola) and by examining contour plots of the fit statistics made by varying the values of two parameters at a time (the contours should be elliptical, and parameter space boundaries should be no closer than approximately 3 sigma from the best-fit point). The int_proj and reg_proj commands may be used for this.

If either of the conditions given above does not hold, then the output from covar may be meaningless except to give an idea of the scale of the confidence intervals. To accurately determine the confidence intervals, one would have to reparameterize the model, use Monte Carlo simulations, or Bayesian methods.

As covar estimates intervals for each parameter independently, the relationship between sigma and the change in statistic value delta_S can be particularly simple: sigma = the square root of delta_S for statistics sampled from the chi-square distribution and for the Cash statistic, and is approximately equal to the square root of (2 * delta_S) for fits based on the general log-likelihood. The default setting is to calculate the one-sigma interval, which can be changed with the sigma option to set_covar_opt or get_covar.

Examples

Evaluate confidence intervals for all thawed parameters in all data sets with an associated source model. The results are then stored in the variable res.

>>> covar()
>>> res = get_covar_results()

Only evaluate the parameters associated with data set 2.

>>> covar(2)

Only evaluate the intervals for the pos.xpos and pos.ypos parameters:

>>> covar(pos.xpos, pos.ypos)

Change the limits to be 1.6 sigma (90%) rather than the default 1 sigma.

>>> set_covar_ope('sigma', 1.6)
>>> covar()

Only evaluate the clus.kt parameter for the data sets with identifiers “obs1”, “obs5”, and “obs6”. This will still use the 1.6 sigma setting from the previous run.

>>> covar("obs1", "obs5", "obs6", clus.kt)

Estimate the errors for all the thawed parameters from the line model and the clus.kt parameter for datasets 1, 3, and 4:

>>> covar(1, 3, 4, line, clus.kt)
covariance(*args) [edit on github]

Estimate parameter confidence intervals using the covariance method.

The covar command computes confidence interval bounds for the specified model parameters in the dataset, using the covariance matrix of the statistic. The get_covar and set_covar_opt commands can be used to configure the error analysis; an example being changing the sigma field to 1.6 (i.e. 90%) from its default value of 1. The output from the routine is displayed on screen, and the get_covar_results routine can be used to retrieve the results.

Parameters
  • id (int or str, optional) – The data set, or sets, that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • parameter (sherpa.models.parameter.Parameter, optional) – The default is to calculate the confidence limits on all thawed parameters of the model, or models, for all the data sets. The evaluation can be restricted by listing the parameters to use. Note that each parameter should be given as a separate argument, rather than as a list. For example covar(g1.ampl, g1.sigma).

  • model (sherpa.models.model.Model, optional) – Select all the thawed parameters in the model.

See also

covar

Estimate the confidence intervals using the confidence method.

get_covar

Return the covariance estimation object.

get_covar_results

Return the results of the last covar run.

int_proj

Plot the statistic value as a single parameter is varied.

int_unc

Plot the statistic value as a single parameter is varied.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

set_covar_opt

Set an option of the covar estimation object.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with multiple ids or parameters values, the order is unimportant, since any argument that is not defined as a model parameter is assumed to be a data id.

The covar command is different to conf, in that in that all other thawed parameters are fixed, rather than being allowed to float to new best-fit values. While conf is more general (e.g. allowing the user to examine the parameter space away from the best-fit point), it is in the strictest sense no more accurate than covar for determining confidence intervals.

An estimated confidence interval is accurate if and only if:

  1. the chi^2 or logL surface in parameter space is approximately shaped like a multi-dimensional paraboloid, and

  2. the best-fit point is sufficiently far from parameter space boundaries.

One may determine if these conditions hold, for example, by plotting the fit statistic as a function of each parameter’s values (the curve should approximate a parabola) and by examining contour plots of the fit statistics made by varying the values of two parameters at a time (the contours should be elliptical, and parameter space boundaries should be no closer than approximately 3 sigma from the best-fit point). The int_proj and reg_proj commands may be used for this.

If either of the conditions given above does not hold, then the output from covar may be meaningless except to give an idea of the scale of the confidence intervals. To accurately determine the confidence intervals, one would have to reparameterize the model, use Monte Carlo simulations, or Bayesian methods.

As covar estimates intervals for each parameter independently, the relationship between sigma and the change in statistic value delta_S can be particularly simple: sigma = the square root of delta_S for statistics sampled from the chi-square distribution and for the Cash statistic, and is approximately equal to the square root of (2 * delta_S) for fits based on the general log-likelihood. The default setting is to calculate the one-sigma interval, which can be changed with the sigma option to set_covar_opt or get_covar.

Examples

Evaluate confidence intervals for all thawed parameters in all data sets with an associated source model. The results are then stored in the variable res.

>>> covar()
>>> res = get_covar_results()

Only evaluate the parameters associated with data set 2.

>>> covar(2)

Only evaluate the intervals for the pos.xpos and pos.ypos parameters:

>>> covar(pos.xpos, pos.ypos)

Change the limits to be 1.6 sigma (90%) rather than the default 1 sigma.

>>> set_covar_ope('sigma', 1.6)
>>> covar()

Only evaluate the clus.kt parameter for the data sets with identifiers “obs1”, “obs5”, and “obs6”. This will still use the 1.6 sigma setting from the previous run.

>>> covar("obs1", "obs5", "obs6", clus.kt)

Estimate the errors for all the thawed parameters from the line model and the clus.kt parameter for datasets 1, 3, and 4:

>>> covar(1, 3, 4, line, clus.kt)
create_model_component(typename=None, name=None)[source] [edit on github]

Create a model component.

Model components created by this function are set to their default values. Components can also be created directly using the syntax typename.name, such as in calls to set_model and set_source (unless you have called set_model_autoassign_func to change the default model auto-assignment setting).

Parameters
  • typename (str) – The name of the model. This should match an entry from the return value of list_models, and defines the type of model.

  • name (str) – The name used to refer to this instance, or component, of the model. A Python variable will be created with this name that can be used to inspect and change the model parameters, as well as use it in model expressions.

Returns

model

Return type

the sherpa.models.Model object created

See also

delete_model_component

Delete a model component.

get_model_component

Returns a model component given its name.

list_models

List the available model types.

list_model_components

List the names of all the model components.

set_model

Set the source model expression for a data set.

set_model_autoassign_func

Set the method used to create model component identifiers.

Notes

This function can over-write an existing component. If the over-written component is part of a source expression - as set by set_model - then the model evaluation will still use the old model definition (and be able to change the fit parameters), but direct access to its parameters is not possible since the name now refers to the new component (this is true using direct access, such as mname.parname, or with set_par).

Examples

Create an instance of the powlaw1d model called pl, and then freeze its gamma parameter to 2.6.

>>> create_model_component("powlaw1d", "pl")
>>> pl.gamma = 2.6
>>> freeze(pl.gamma)

Create a blackbody model called bb, check that it is reconized as a component, and display its parameters:

>>> create_model_component("bbody", "bb")
>>> list_model_components()
>>> print(bb)
>>> print(bb.ampl)
dataspace1d(start, stop, step=1, numbins=None, id=None, dstype=<class 'sherpa.data.Data1DInt'>)[source] [edit on github]

Create the independent axis for a 1D data set.

Create an “empty” one-dimensional data set by defining the grid on which the points are defined (the independent axis). The values are set to 0.

Parameters
  • start (number) – The minimum value of the axis.

  • stop (number) – The maximum value of the axis.

  • step (number, optional) – The separation between each grid point. This is not used if numbins is set.

  • numbins (int, optional) – The number of grid points. This over-rides the step setting.

  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • dstype (data class to use, optional) – What type of data is to be used. Supported values include Data1DInt (the default) and Data1D.

See also

dataspace2d

Create the independent axis for a 2D data set.

get_dep

Return the dependent axis of a data set.

get_indep

Return the independent axes of a data set.

set_dep

Set the dependent axis of a data set.

Notes

The meaning of the stop parameter depends on whether it is a binned or unbinned data set (as set by the dstype parameter).

Examples

Create a binned data set, starting at 1 and with a bin-width of 1.

>>> dataspace1d(1, 5, 1)
>>> print(get_indep())
(array([ 1.,  2.,  3.,  4.]), array([ 2.,  3.,  4.,  5.]))

This time for an un-binned data set:

>>> dataspace1d(1, 5, 1, dstype=Data1D)
>>> print(get_indep())
(array([ 1.,  2.,  3.,  4.,  5.]),)

Specify the number of bins rather than the grid spacing:

>>> dataspace1d(1, 5, numbins=5, id=2)
>>> (xlo, xhi) = get_indep(2)
>>> xlo
array([ 1. ,  1.8,  2.6,  3.4,  4.2])
>>> xhi
array([ 1.8,  2.6,  3.4,  4.2,  5. ])
>>> dataspace1d(1, 5, numbins=5, id=3, dstype=Data1D)
>>> (x, ) = get_indep(3)
>>> x
array([ 1.,  2.,  3.,  4.,  5.])
dataspace2d(dims, id=None, dstype=<class 'sherpa.data.Data2D'>)[source] [edit on github]

Create the independent axis for a 2D data set.

Create an “empty” two-dimensional data set by defining the grid on which the points are defined (the independent axis). The values are set to 0.

Parameters
  • dims (sequence of 2 number) – The dimensions of the grid in (width,height) order.

  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • dstype (data class to use, optional) – What type of data is to be used. Supported values include Data2D (the default) and Data2DInt.

See also

dataspace1d

Create the independent axis for a 1D data set.

get_dep

Return the dependent axis of a data set.

get_indep

Return the independent axes of a data set.

set_dep

Set the dependent axis of a data set.

Examples

Create a 200 pixel by 150 pixel grid (number of columns by number of rows) and display it (each pixel has a value of 0):

>>> dataspace2d([200, 150])
>>> image_data()

Create a data space called “fakeimg”:

>>> dataspace2d([nx,ny], id="fakeimg")
delete_data(id=None)[source] [edit on github]

Delete a data set by identifier.

The data set, and any associated structures - such as the ARF and RMF for PHA data sets - are removed.

Parameters

id (int or str, optional) – The data set to delete. If not given then the default identifier is used, as returned by get_default_id.

See also

clean

Clear all stored session data.

copy_data

Copy a data set to a new identifier.

delete_model

Delete the model expression from a data set.

get_default_id

Return the default data set identifier.

list_data_ids

List the identifiers for the loaded data sets.

Notes

The source expression is not removed by this function.

Examples

Delete the data from the default data set:

>>> delete_data()

Delete the data set identified as ‘src’:

>>> delete_data('src')
delete_model(id=None)[source] [edit on github]

Delete the model expression for a data set.

This removes the model expression, created by set_model, for a data set. It does not delete the components of the expression.

Parameters

id (int or str, optional) – The data set containing the source expression. If not given then the default identifier is used, as returned by get_default_id.

See also

clean

Clear all stored session data.

delete_data

Delete a data set by identifier.

get_default_id

Return the default data set identifier.

set_model

Set the source model expression for a data set.

show_model

Display the source model expression for a data set.

Examples

Remove the model expression for the default data set:

>>> delete_model()

Remove the model expression for the data set with the identifier called ‘src’:

>>> delete_model('src')
delete_model_component(name)[source] [edit on github]

Delete a model component.

Parameters

name (str) – The name used to refer to this instance, or component, of the model. The corresponding Python variable will be deleted by this function.

See also

create_model_component

Create a model component.

delete_model

Delete the model expression for a data set.

list_models

List the available model types.

list_model_components

List the names of all the model components.

set_model

Set the source model expression for a data set.

set_model_autoassign_func

Set the method used to create model component identifiers.

Notes

It is an error to try to delete a component that is part of a model expression - i.e. included as part of an expression in a set_model or set_source call. In such a situation, use the delete_model function to remove the source expression before calling delete_model_component.

Examples

If a model instance called pl has been created - e.g. by create_model_component('powlaw1d', 'pl') - then the following will remove it:

>>> delete_model_component('pl')
delete_psf(id=None)[source] [edit on github]

Delete the PSF model for a data set.

Remove the PSF convolution applied to a source model.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

See also

list_psf_ids

List of all the data sets with a PSF.

load_psf

Create a PSF model.

set_psf

Add a PSF model to a data set.

get_psf

Return the PSF model defined for a data set.

Examples

>>> delete_psf()
>>> delete_psf('core')
fake(id=None, method=<function poisson_noise>)[source] [edit on github]

Simulate a data set.

Take a data set, evaluate the model for each bin, and then use this value to create a data value from each bin. The default behavior is to use a Poisson distribution, with the model value as the expectation value of the distribution.

Parameters
  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • method (func) – The function used to create a random realisation of a data set.

See also

dataspace1d

Create the independent axis for a 1D data set.

dataspace2d

Create the independent axis for a 2D data set.

get_dep

Return the dependent axis of a data set.

load_arrays

Create a data set from array values.

set_model

Set the source model expression for a data set.

Notes

The function for the method argument accepts a single argument, the data values, and should return an array of the same shape as the input, with the data values to use.

The function can be called on any data set, it does not need to have been created with dataspace1d or dataspace2d.

Specific data set types may have their own, specialized, version of this function.

Examples

Create a random realisation of the model - a constant plus gaussian line - for the range x=-5 to 5.

>>> dataspace1d(-5, 5, 0.5, dstype=Data1D)
>>> set_source(gauss1d.gline + const1d.bgnd)
>>> bgnd.c0 = 2
>>> gline.fwhm = 4
>>> gline.ampl = 5
>>> gline.pos = 1
>>> fake()
>>> plot_data()
>>> plot_model(overplot=True)

For a 2D data set, display the simulated data, model, and residuals:

>>> dataspace2d([150, 80], id='fakeimg')
>>> set_source('fakeimg', beta2d.src + polynom2d.bg)
>>> src.xpos, src.ypos = 75, 40
>>> src.r0, src.alpha = 15, 2.3
>>> src.ellip, src.theta = 0.4, 1.32
>>> src.ampl = 100
>>> bg.c, bg.cx1, bg.cy1 = 3, 0.4, 0.3
>>> fake('fakeimg')
>>> image_fit('fakeimg')
fit(id=None, *otherids, **kwargs)[source] [edit on github]

Fit a model to one or more data sets.

Use forward fitting to find the best-fit model to one or more data sets, given the chosen statistic and optimization method. The fit proceeds until the results converge or the number of iterations exceeds the maximum value (these values can be changed with set_method_opt). An iterative scheme can be added using set_iter_method to try and improve the fit. The final fit results are displayed to the screen and can be retrieved with get_fit_results.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then all data sets with an associated model are fit simultaneously.

  • *otherids (int or str, optional) – Other data sets to use in the calculation.

  • outfile (str, optional) – If set, then the fit results will be written to a file with this name. The file contains the per-iteration fit results.

  • clobber (bool, optional) – This flag controls whether an existing file can be overwritten (True) or if it raises an exception (False, the default setting).

Raises

sherpa.utils.err.FitErr – If filename already exists and clobber is False.

See also

conf

Estimate parameter confidence intervals using the confidence method.

contour_fit

Contour the fit to a data set.

covar

Estimate the confidence intervals using the confidence method.

freeze

Fix model parameters so they are not changed by a fit.

get_fit_results

Return the results of the last fit.

plot_fit

Plot the fit results (data, model) for a data set.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

set_stat

Set the statistical method.

set_method

Change the optimization method.

set_method_opt

Change an option of the current optimization method.

set_full_model

Define the convolved model expression for a data set.

set_iter_method

Set the iterative-fitting scheme used in the fit.

set_model

Set the model expression for a data set.

show_fit

Summarize the fit results.

thaw

Allow model parameters to be varied during a fit.

Examples

Simultaneously fit all data sets with models and then store the results in the variable fres:

>>> fit()
>>> fres = get_fit_results()

Fit just the data set ‘img’:

>>> fit('img')

Simultaneously fit data sets 1, 2, and 3:

>>> fit(1, 2, 3)

Fit data set ‘jet’ and write the fit results to the text file ‘jet.fit’, over-writing it if it already exists:

>>> fit('jet', outfile='jet.fit', clobber=True)
freeze(*args)[source] [edit on github]

Fix model parameters so they are not changed by a fit.

The arguments can be parameters or models, in which case all parameters of the model are frozen. If no arguments are given then nothing is changed.

Parameters

args (sequence of str or Parameter or Model) – The parameters or models to freeze.

See also

fit

Fit one or more data sets.

link

Link a parameter value to an associated value.

set_par

Set the value, limits, or behavior of a model parameter.

thaw

Allow model parameters to be varied during a fit.

unlink

Unlink a parameter value.

Notes

The thaw function can be used to reverse this setting, so that parameters can be varied in a fit.

Examples

Fix the FWHM parameter of the line model (in this case a gauss1d model) so that it will not be varied in the fit.

>>> set_source(const1d.bgnd + gauss1d.line)
>>> line.fwhm = 2.1
>>> freeze(line.fwhm)
>>> fit()

Freeze all parameters of the line model and then re-fit:

>>> freeze(line)
>>> fit()

Freeze the nh parameter of the gal model and the abund parameter of the src model:

>>> freeze(gal.nh, src.abund)
get_cdf_plot()[source] [edit on github]

Return the data used to plot the last CDF.

Returns

plot – An object containing the data used by the last call to plot_cdf. The fields will be None if the function has not been called.

Return type

a sherpa.plot.CDFPlot instance

See also

plot_cdf

Plot the cumulative density function of an array.

get_chisqr_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_chisqr.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_chisqr (or get_chisqr_plot) are returned, otherwise the data is re-generated.

Returns

resid_data

Return type

a sherpa.plot.ChisqrPlot instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_delchi_plot

Return the data used by plot_delchi.

get_ratio_plot

Return the data used by plot_ratio.

get_resid_plot

Return the data used by plot_resid.

plot_chisqr

Plot the chi-squared value for each point in a data set.

Examples

Return the residual data, measured as chi square, for the default data set:

>>> rplot = get_chisqr_plot()
>>> np.min(rplot.y)
0.0005140622701128954
>>> np.max(rplot.y)
8.379696454792295

Display the contents of the residuals plot for data set 2:

>>> print(get_chisqr_plot(2))

Overplot the residuals plot from the ‘core’ data set on the ‘jet’ data set:

>>> r1 = get_chisqr_plot('jet')
>>> r2 = get_chisqr_plot('core')
>>> r1.plot()
>>> r2.overplot()
get_conf()[source] [edit on github]

Return the confidence-interval estimation object.

Returns

conf

Return type

object

See also

conf

Estimate parameter confidence intervals using the confidence method.

get_conf_opt

Return one or all of the options for the confidence interval method.

set_conf_opt

Set an option of the conf estimation object.

Notes

The attributes of the confidence-interval object include:

eps

The precision of the calculated limits. The default is 0.01.

fast

If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

max_rstat

If the reduced chi square is larger than this value, do not use (only used with chi-square statistics). The default is 3.

maxfits

The maximum number of re-fits allowed (that is, when the remin filter is met). The default is 5.

maxiters

The maximum number of iterations allowed when bracketing limits, before stopping for that parameter. The default is 200.

numcores

The number of computer cores to use when evaluating results in parallel. This is only used if parallel is True. The default is to use all cores.

openinterval

How the conf method should cope with intervals that do not converge (that is, when the maxiters limit has been reached). The default is False.

parallel

If there is more than one free parameter then the results can be evaluated in parallel, to reduce the time required. The default is True.

remin

The minimum difference in statistic value for a new fit location to be considered better than the current best fit (which starts out as the starting location of the fit at the time conf is called). The default is 0.01.

sigma

What is the error limit being calculated. The default is 1.

soft_limits

Should the search be restricted to the soft limits of the parameters (True), or can parameter values go out all the way to the hard limits if necessary (False). The default is False

tol

The tolerance for the fit. The default is 0.2.

verbose

Should extra information be displayed during fitting? The default is False.

Examples

>>> print(get_conf())
name         = confidence
numcores     = 8
verbose      = False
openinterval = False
max_rstat    = 3
maxiters     = 200
soft_limits  = False
eps          = 0.01
fast         = False
maxfits      = 5
remin        = 0.01
tol          = 0.2
sigma        = 1
parallel     = True

Change the remin field to 0.05.

>>> cf = get_conf()
>>> cf.remin = 0.05
get_conf_opt(name=None)[source] [edit on github]

Return one or all of the options for the confidence interval method.

This is a helper function since the options can also be read directly using the object returned by get_conf.

Parameters

name (str, optional) – If not given, a dictionary of all the options are returned. When given, the individual value is returned.

Returns

value

Return type

dictionary or value

Raises

sherpa.utils.err.ArgumentErr – If the name argument is not recognized.

See also

conf

Estimate parameter confidence intervals using the confidence method.

get_conf

Return the confidence-interval estimation object.

set_conf_opt

Set an option of the conf estimation object.

Examples

>>> get_conf_opt('verbose')
False
>>> copts = get_conf_opt()
>>> copts['verbose']
False
get_conf_results()[source] [edit on github]

Return the results of the last conf run.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no conf call has been made.

See also

get_conf_opt

Return one or all of the options for the confidence interval method.

set_conf_opt

Set an option of the conf estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘confidence’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

Examples

>>> res = get_conf_results()
>>> print(res)
datasets    = (1,)
methodname  = confidence
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('p1.gamma', 'p1.ampl')
parvals     = (2.1585155113403327, 0.00022484014787994827)
parmins     = (-0.082785567348122591, -1.4825550342799376e-05)
parmaxes    = (0.083410634144100104, 1.4825550342799376e-05)
nfits       = 13

The following converts the above into a dictionary where the keys are the parameter names and the values are the tuple (best-fit value, lower-limit, upper-limit):

>>> pvals1 = zip(res.parvals, res.parmins, res.parmaxes)
>>> pvals2 = [(v, v+l, v+h) for (v, l, h) in pvals1]
>>> dres = dict(zip(res.parnames, pvals2))
>>> dres['p1.gamma']
(2.1585155113403327, 2.07572994399221, 2.241926145484433)
get_confidence_results() [edit on github]

Return the results of the last conf run.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no conf call has been made.

See also

get_conf_opt

Return one or all of the options for the confidence interval method.

set_conf_opt

Set an option of the conf estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘confidence’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

Examples

>>> res = get_conf_results()
>>> print(res)
datasets    = (1,)
methodname  = confidence
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('p1.gamma', 'p1.ampl')
parvals     = (2.1585155113403327, 0.00022484014787994827)
parmins     = (-0.082785567348122591, -1.4825550342799376e-05)
parmaxes    = (0.083410634144100104, 1.4825550342799376e-05)
nfits       = 13

The following converts the above into a dictionary where the keys are the parameter names and the values are the tuple (best-fit value, lower-limit, upper-limit):

>>> pvals1 = zip(res.parvals, res.parmins, res.parmaxes)
>>> pvals2 = [(v, v+l, v+h) for (v, l, h) in pvals1]
>>> dres = dict(zip(res.parnames, pvals2))
>>> dres['p1.gamma']
(2.1585155113403327, 2.07572994399221, 2.241926145484433)
get_covar()[source] [edit on github]

Return the covariance estimation object.

Returns

covar

Return type

object

See also

covar

Estimate parameter confidence intervals using the covariance method.

get_covar_opt

Return one or all of the options for the covariance method.

set_covar_opt

Set an option of the covar estimation object.

Notes

The attributes of the covariance object include:

eps

The precision of the calculated limits. The default is 0.01.

maxiters

The maximum number of iterations allowed before stopping for that parameter. The default is 200.

sigma

What is the error limit being calculated. The default is 1.

soft_limits

Should the search be restricted to the soft limits of the parameters (True), or can parameter values go out all the way to the hard limits if necessary (False). The default is False

Examples

>>> print(get_covar())
name        = covariance
sigma       = 1
maxiters    = 200
soft_limits = False
eps         = 0.01

Change the sigma field to 1.9.

>>> cv = get_covar()
>>> cv.sigma = 1.6
get_covar_opt(name=None)[source] [edit on github]

Return one or all of the options for the covariance method.

This is a helper function since the options can also be read directly using the object returned by get_covar.

Parameters

name (str, optional) – If not given, a dictionary of all the options are returned. When given, the individual value is returned.

Returns

value

Return type

dictionary or value

Raises

sherpa.utils.err.ArgumentErr – If the name argument is not recognized.

See also

covar

Estimate parameter confidence intervals using the covariance method.

get_covar

Return the covariance estimation object.

set_covar_opt

Set an option of the covar estimation object.

Examples

>>> get_covar_opt('sigma')
1
>>> copts = get_covar_opt()
>>> copts['sigma']
1
get_covar_results()[source] [edit on github]

Return the results of the last covar run.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no covar call has been made.

See also

get_covar_opt

Return one or all of the options for the covariance method.

set_covar_opt

Set an option of the covar estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘covariance’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

There is also an extra_output field which is used to return the covariance matrix.

Examples

>>> res = get_covar_results()
>>> print(res)
datasets    = (1,)
methodname  = covariance
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('bgnd.c0',)
parvals     = (10.228675427602724,)
parmins     = (-2.4896739438296795,)
parmaxes    = (2.4896739438296795,)
nfits       = 0

In this case, of a single parameter, the covariance matrix is just the variance of the parameter:

>>> res.extra_output
array([[ 6.19847635]])
get_covariance_results() [edit on github]

Return the results of the last covar run.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no covar call has been made.

See also

get_covar_opt

Return one or all of the options for the covariance method.

set_covar_opt

Set an option of the covar estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘covariance’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

There is also an extra_output field which is used to return the covariance matrix.

Examples

>>> res = get_covar_results()
>>> print(res)
datasets    = (1,)
methodname  = covariance
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('bgnd.c0',)
parvals     = (10.228675427602724,)
parmins     = (-2.4896739438296795,)
parmaxes    = (2.4896739438296795,)
nfits       = 0

In this case, of a single parameter, the covariance matrix is just the variance of the parameter:

>>> res.extra_output
array([[ 6.19847635]])
get_data(id=None)[source] [edit on github]

Return the data set by identifier.

The object returned by the call can be used to query and change properties of the data set.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

An instance of a sherpa.Data.Data-derived class.

Return type

instance

Raises

sherpa.utils.err.IdentifierErr – If no model expression has been set for the data set (with set_model or set_source).

See also

copy_data

Copy a data set to a new identifier.

delete_data

Delete a data set by identifier.

load_data

Create a data set from a file.

set_data

Set a data set.

Examples

>>> d = get_data()
>>> dimg = get_data('img')
>>> load_arrays('tst', [10, 20, 30], [5.4, 2.3, 9.8])
>>> print(get_data('tst'))
name      =
x         = Int64[3]
y         = Float64[3]
staterror = None
syserror  = None
get_data_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_data.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_data (or get_data_contour) are returned, otherwise the data is re-generated.

Returns

resid_data – The y attribute contains the residual values and the x0 and x1 arrays contain the corresponding coordinate values, as one-dimensional arrays.

Return type

a sherpa.plot.DataContour instance

Raises

See also

get_data_image

Return the data used by image_data.

contour_data

Contour the values of an image data set.

image_data

Display a data set in the image viewer.

Examples

Return the data for the default data set:

>>> dinfo = get_data_contour()
get_data_contour_prefs()[source] [edit on github]

Return the preferences for contour_data.

Returns

prefs – Changing the values of this dictionary will change any new contour plots. The default is an empty dictionary.

Return type

dict

See also

contour_data

Contour the values of an image data set.

Notes

The meaning of the fields depend on the chosen plot backend. A value of None (or not set) means to use the default value for that attribute, unless indicated otherwise.

alpha

The transparency value used to draw the contours, where 0 is fully transparent and 1 is fully opaque.

colors

The colors to draw the contours.

linewidths

What thickness of line to draw the contours.

xlog

Should the X axis be drawn with a logarithmic scale? The default is False.

ylog

Should the Y axis be drawn with a logarithmic scale? The default is False.

Examples

Change the contours to be drawn in ‘green’:

>>> contour_data()
>>> prefs = get_data_contour_prefs()
>>> prefs['color'] = 'green'
>>> contour_data()
get_data_image(id=None)[source] [edit on github]

Return the data used by image_data.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

data_img – The y attribute contains the ratio values as a 2D NumPy array.

Return type

a sherpa.image.DataImage instance

Raises

See also

contour_data

Contour the values of an image data set.

image_data

Display a data set in the image viewer.

Examples

Return the image data for the default data set:

>>> dinfo = get_data_image()
>>> dinfo.y.shape
(150, 175)
get_data_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_data.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_data (or get_data_plot) are returned, otherwise the data is re-generated.

Returns

data – An object representing the data used to create the plot by plot_data. The relationship between the returned values and the values in the data set depend on the data type. For example PHA data are plotted in units controlled by sherpa.astro.ui.set_analysis, but are stored as channels and counts, and may have been grouped and the background estimate removed.

Return type

a sherpa.plot.DataPlot instance

See also

get_data_plot_prefs

Return the preferences for plot_data.

get_default_id

Return the default data set identifier.

plot_data

Plot the data values.

get_data_plot_prefs(id=None)[source] [edit on github]

Return the preferences for plot_data.

The plot preferences may depend on the data set, so it is now an optional argument.

Changed in version 4.12.2: The id argument has been given.

Parameters

id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

Returns

prefs – Changing the values of this dictionary will change any new data plots. This dictionary will be empty if no plot backend is available.

Return type

dict

See also

plot_data

Plot the data values.

set_xlinear

New plots will display a linear X axis.

set_xlog

New plots will display a logarithmically-scaled X axis.

set_ylinear

New plots will display a linear Y axis.

set_ylog

New plots will display a logarithmically-scaled Y axis.

Notes

The meaning of the fields depend on the chosen plot backend. A value of None means to use the default value for that attribute, unless indicated otherwise. These preferences are used by the following commands: plot_data, plot_bkg, plot_ratio, and the “fit” variants, such as plot_fit, plot_fit_resid, and plot_bkg_fit.

The following preferences are recognized by the matplotlib backend:

alpha

The transparency value used to draw the line or symbol, where 0 is fully transparent and 1 is fully opaque.

barsabove

The barsabove argument for the matplotlib errorbar function.

capsize

The capsize argument for the matplotlib errorbar function.

color

The color to use (will be over-ridden by more-specific options below). The default is None.

ecolor

The color to draw error bars. The default is None.

linestyle

How should the line connecting the data points be drawn. The default is ‘None’, which means no line is drawn.

marker

What style is used for the symbols. The default is '.' which indicates a point.

markerfacecolor

What color to draw the symbol representing the data points. The default is None.

markersize

What size is the symbol drawn. The default is None.

ratioline

Should a horizontal line be drawn at y=1? The default is False.

xaxis

The default is False

xerrorbars

Should error bars be drawn for the X axis. The default is False.

xlog

Should the X axis be drawn with a logarithmic scale? The default is False. This field can also be changed with the set_xlog and set_xlinear functions.

yerrorbars

Should error bars be drawn for the Y axis. The default is True.

ylog

Should the Y axis be drawn with a logarithmic scale? The default is False. This field can also be changed with the set_ylog and set_ylinear functions.

Examples

After these commands, any data plot will use a green symbol and not display Y error bars.

>>> prefs = get_data_plot_prefs()
>>> prefs['color'] = 'green'
>>> prefs['yerrorbars'] = False
get_default_id()[source] [edit on github]

Return the default data set identifier.

The Sherpa data id ties data, model, fit, and plotting information into a data set easily referenced by id. The default identifier, used by many commands, is returned by this command and can be changed by set_default_id.

Returns

id – The default data set identifier used by certain Sherpa functions when an identifier is not given, or set to None.

Return type

int or str

See also

list_data_ids

List the identifiers for the loaded data sets.

set_default_id

Set the default data set identifier.

Notes

The default Sherpa data set identifier is the integer 1.

Examples

Display the default identifier:

>>> print(get_default_id())

Store the default identifier and use it as an argument to call another Sherpa routine:

>>> defid = get_default_id()
>>> load_arrays(defid, x, y)
get_delchi_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_delchi.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_delchi (or get_delchi_plot) are returned, otherwise the data is re-generated.

Returns

resid_data

Return type

a sherpa.plot.DelchiPlot instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_chisqr_plot

Return the data used by plot_chisqr.

get_ratio_plot

Return the data used by plot_ratio.

get_resid_plot

Return the data used by plot_resid.

plot_delchi

Plot the ratio of residuals to error for a data set.

Examples

Return the residual data, measured in units of the error, for the default data set:

>>> rplot = get_delchi_plot()
>>> np.min(rplot.y)
-2.85648373819671875
>>> np.max(rplot.y)
2.89477053577520982

Display the contents of the residuals plot for data set 2:

>>> print(get_delchi_plot(2))

Overplot the residuals plot from the ‘core’ data set on the ‘jet’ data set:

>>> r1 = get_delchi_plot('jet')
>>> r2 = get_delchi_plot('core')
>>> r1.plot()
>>> r2.overplot()
get_dep(id=None, filter=False)[source] [edit on github]

Return the dependent axis of a data set.

This function returns the data values (the dependent axis) for each point or pixel in the data set.

Parameters
  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • filter (bool, optional) – Should the filter attached to the data set be applied to the return value or not. The default is False.

Returns

axis – The dependent axis values. The model estimate is compared to these values during fitting.

Return type

array

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_error

Return the errors on the dependent axis of a data set.

get_indep

Return the independent axis of a data set.

list_data_ids

List the identifiers for the loaded data sets.

Examples

>>> load_arrays(1, [10, 15, 19], [4, 5, 9])
>>> get_dep()
array([4, 5, 9])
>>> x0 = [10, 15, 12, 19]
>>> x1 = [12, 14, 10, 17]
>>> y = [4, 5, 9, -2]
>>> load_arrays(2, x0, x1, y, Data2D)
>>> get_dep(2)
array([ 4,  5,  9, -2])

If the filter flag is set then the return will be limited to the data that is used in the fit:

>>> load_arrays(1, [10, 15, 19], [4, 5, 9])
>>> ignore_id(1, 17, None)
>>> get_dep()
array([4, 5, 9])
>>> get_dep(filter=True)
array([4, 5])
get_dims(id=None, filter=False)[source] [edit on github]

Return the dimensions of the data set.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • filter (bool, optional) – If True then apply any filter to the data set before returning the dimensions. The default is False.

Returns

dims

Return type

a tuple of int

See also

ignore

Exclude data from the fit.

sherpa.astro.ui.ignore2d

Exclude a spatial region from an image.

notice

Include data in the fit.

sherpa.astro.ui.notice2d

Include a spatial region of an image.

Examples

Display the dimensions for the default data set:

>>> print(get_dims())

Find the number of bins in dataset ‘a2543’ without and with any filters applied to it:

>>> nall = get_dims('a2543')
>>> nfilt = get_dims('a2543', filter=True)
get_draws(id=None, otherids=(), niter=1000, covar_matrix=None)[source] [edit on github]

Run the pyBLoCXS MCMC algorithm.

The function runs a Markov Chain Monte Carlo (MCMC) algorithm designed to carry out Bayesian Low-Count X-ray Spectral (BLoCXS) analysis. It explores the model parameter space at the suspected statistic minimum (i.e. after using fit). The return values include the statistic value, parameter values, and an acceptance flag indicating whether the row represents a jump from the current location or not. For more information see the sherpa.sim module and [1]_.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (sequence of int or str, optional) – Other data sets to use in the calculation.

  • niter (int, optional) – The number of draws to use. The default is 1000.

  • covar_matrix (2D array, optional) – The covariance matrix to use. If None then the result from get_covar_results().extra_output is used.

Returns

The results of the MCMC chain. The stats and accept arrays contain niter+1 elements, with the first row being the starting values. The params array has (nparams, niter+1) elements, where nparams is the number of free parameters in the model expression, and the first column contains the values that the chain starts at. The accept array contains boolean values, indicating whether the jump, or step, was accepted (True), so the parameter values and statistic change, or it wasn’t, in which case there is no change to the previous row. The sherpa.utils.get_error_estimates routine can be used to calculate the credible one-sigma interval from the params array.

Return type

stats, accept, params

See also

covar

Estimate the confidence intervals using the covariance method.

fit

Fit a model to one or more data sets.

plot_cdf

Plot the cumulative density function of an array.

plot_pdf

Plot the probability density function of an array.

plot_scatter

Create a scatter plot.

plot_trace

Create a trace plot of row number versus value.

set_prior

Set the prior function to use with a parameter.

set_sampler

Set the MCMC sampler.

get_sampler

Return information about the current MCMC sampler.

Notes

The chain is run using fit information associated with the specified data set, or sets, the currently set sampler (set_sampler) and parameter priors (set_prior), for a specified number of iterations. The model should have been fit to find the best-fit parameters, and covar called, before running get_draws. The results from get_draws is used to estimate the parameter distributions.

References

1

“Analysis of Energy Spectra with Low Photon Counts via Bayesian Posterior Simulation”, van Dyk, D.A., Connors, A., Kashyap, V.L., & Siemiginowska, A. 2001, Ap.J., 548, 224 http://adsabs.harvard.edu/abs/2001ApJ…548..224V

Examples

Fit a source and then run a chain to investigate the parameter distributions. The distribution of the stats values created by the chain is then displayed, using plot_trace, and the parameter distributions for the first two thawed parameters are displayed. The first one as a cumulative distribution using plot_cdf and the second one as a probability distribution using plot_pdf. Finally the acceptance fraction (number of draws where the chain moved) is displayed. Note that in a full analysis session a burn-in period would normally be removed from the chain before using the results.

>>> fit()
>>> covar()
>>> stats, accept, params = get_draws(1, niter=1e4)
>>> plot_trace(stats, name='stat')
>>> names = [p.fullname for p in get_source().pars if not p.frozen]
>>> plot_cdf(params[0,:], name=names[0], xlabel=names[0])
>>> plot_pdf(params[1,:], name=names[1], xlabel=names[1])
>>> accept[:-1].sum() * 1.0 / len(accept - 1)
0.4287

The following runs the chain on multiple data sets, with identifiers ‘core’, ‘jet1’, and ‘jet2’:

>>> stats, accept, params = get_draws('core', ['jet1', 'jet2'], niter=1e4)
get_error(id=None, filter=False)[source] [edit on github]

Return the errors on the dependent axis of a data set.

The function returns the total errors (a quadrature addition of the statistical and systematic errors) on the values (dependent axis) of a data set. The individual components can be retrieved with the get_staterror and get_syserror functions.

Parameters
  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • filter (bool, optional) – Should the filter attached to the data set be applied to the return value or not. The default is False.

Returns

errors – The error for each data point, formed by adding the statistical and systematic errors in quadrature. The size of this array depends on the filter argument.

Return type

array

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_error

Return the errors on the dependent axis of a data set.

get_indep

Return the independent axis of a data set.

get_staterror

Return the statistical errors on the dependent axis of a data set.

get_syserror

Return the systematic errors on the dependent axis of a data set.

list_data_ids

List the identifiers for the loaded data sets.

Notes

The default behavior is to not apply any filter defined on the independent axes to the results, so that the return value is for all points (or bins) in the data set. Set the filter argument to True to apply this filter.

Examples

Return the error values for the default data set, ignoring any filter applied to it:

>>> err = get_error()

Ensure that the return values are for the selected (filtered) points in the default data set (the return array may be smaller than in the previous example):

>>> err = get_error(filter=True)

Find the errors for the “core” data set:

>>> err = get_error('core', filter=True)
get_filter(id=None)[source] [edit on github]

Return the filter expression for a data set.

This returns the filter expression, created by one or more calls to ignore and notice, for the data set.

Changed in version 4.14.0: The filter expressions have been tweaked for Data1DInt and PHA data sets (when using energy or wavelength units) and now describe the full range of the bins, rather than the mid-points.

Parameters

id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

Returns

filter – The empty string or a string expression representing the filter used. For PHA data dets the units are controlled by the analysis setting for the data set.

Return type

str

Raises

sherpa.utils.err.ArgumentErr – If the data set does not exist.

See also

ignore

Exclude data from the fit.

load_filter

Load the filter array from a file and add to a data set.

notice

Include data in the fit.

save_filter

Save the filter array to a file.

show_filter

Show any filters applied to a data set.

set_filter

Set the filter array of a data set.

Examples

The default filter is the full dataset, given in the format lowval:hival (for a Data1D dataset like this these are inclusive limits):

>>> load_arrays(1, [10, 15, 20, 25], [5, 7, 4, 2])
>>> get_filter()
'10.0000:25.0000'

The notice call restricts the data to the range between 14 and 30. The resulting filter is the combination of this range and the data:

>>> notice(14, 30)
>>> get_filter()
'15.0000:25.0000'

Ignoring the point at x=20 means that only the points at x=15 and x=25 remain, so a comma-separated list is used:

>>> ignore(19, 22)
>>> get_filter()
'15.0000,25.0000'

The filter equivalent to the per-bin array of filter values:

>>> set_filter([1, 1, 0, 1])
>>> get_filter()
'10.0000:15.0000,25.0000'

For an integrated data set (Data1DInt and DataPHA with energy or wavelength units)

>>> load_arrays(1, [10, 15, 20, 25], [15, 20, 23, 30], [5, 7, 4, 2], Data1DInt)
>>> get_filter()
'10.0000:30.0000'

For integrated datasets the limits are now inclusive only for the lower limit, but in this the end-point ends within a bin so is is included:

>>> notice(17, 28)
>>> get_filter()
'15.0000:30.0000'

There is no data in the range 23 to 24 so the ignore doesn’t change anything:

>>> ignore(23, 24)
>>> get_filter()
'15.0000:30.0000'

However it does match the range 22 to 23 and so changes the filter:

>>> ignore(22, 23)
>>> get_filter()
'15.0000:20.0000,25:000:30.0000'

Return the filter for data set 3:

>>> get_filter(3)
get_fit_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_fit.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_fit (or get_fit_contour) are returned, otherwise the data is re-generated.

Returns

fit_data – An object representing the data used to create the plot by contour_fit. It contains the data from get_data_contour and get_model_contour in the datacontour and modelcontour attributes.

Return type

a sherpa.plot.FitContour instance

Raises

See also

get_data_image

Return the data used by image_data.

get_model_image

Return the data used by image_model.

contour_data

Contour the values of an image data set.

contour_model

Contour the values of the model, including any PSF.

image_data

Display a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

Examples

Return the contour data for the default data set:

>>> finfo = get_fit_contour()
get_fit_plot(id=None, recalc=True)[source] [edit on github]

Return the data used to create the fit plot.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_fit (or get_fit_plot) are returned, otherwise the data is re-generated.

Returns

data – An object representing the data used to create the plot by plot_fit. It contains the data from get_data_plot and get_model_plot in the dataplot and modelplot attributes.

Return type

a sherpa.plot.FitPlot instance

See also

get_data_plot_prefs

Return the preferences for plot_data.

get_model_plot_prefs

Return the preferences for plot_model.

get_default_id

Return the default data set identifier.

plot_data

Plot the data values.

plot_model

Plot the model for a data set.

Examples

Create the data needed to create the “fit plot” for the default data set and display it:

>>> fplot = get_fit_plot()
>>> print(fplot)

Return the plot data for data set 2, and then use it to create a plot:

>>> f2 = get_fit_plot(2)
>>> f2.plot()

The fit plot consists of a combination of a data plot and a model plot, which are captured in the dataplot and modelplot attributes of the return value. These can be used to display the plots individually, such as:

>>> f2.dataplot.plot()
>>> f2.modelplot.plot()

or, to combine the two:

>>> f2.dataplot.plot()
>>> f2.modelplot.overplot()
get_fit_results()[source] [edit on github]

Return the results of the last fit.

This function returns the results from the most-recent fit. The returned value includes information on the parameter values and fit statistic.

Returns

stats – The results of the last fit. It does not reflect any changes made to the model parameter, or settings, since the last fit.

Return type

a sherpa.fit.FitResults instance

See also

calc_stat

Calculate the fit statistic for a data set.

calc_stat_info

Display the statistic values for the current models.

fit

Fit a model to one or more data sets.

get_stat_info

Return the statistic values for the current models.

set_iter_method

Set the iterative-fitting scheme used in the fit.

Notes

The fields of the object include:

datasets

A sequence of the data set ids included in the results.

itermethodname

What iterated-fit scheme was used, if any (as set by set_iter_method).

statname

The name of the statistic function (as used in set_stat).

succeeded

Was the fit successful (did it converge)?

parnames

A tuple of the parameter names that were varied in the fit (the thawed parameters in the model expression).

parvals

A tuple of the parameter values, in the same order as parnames.

statval

The statistic value after the fit.

istatval

The statistic value at the start of the fit.

dstatval

The change in the statistic value (istatval - statval).

numpoints

The number of bins used in the fits.

dof

The number of degrees of freedom in the fit (the number of bins minus the number of free parameters).

qval

The Q-value (probability) that one would observe the reduced statistic value, or a larger value, if the assumed model is true and the current model parameters are the true parameter values. This will be None if the value can not be calculated with the current statistic (e.g. the Cash statistic).

rstat

The reduced statistic value (the statval field divided by dof). This is not calculated for all statistics.

message

A message about the results of the fit (e.g. if the fit was unable to converge). The format and contents depend on the optimisation method.

nfev

The number of model evaluations made during the fit.

Examples

Display the fit results:

>>> print(get_fit_results())

Inspect the fit results:

>>> res = get_fit_results()
>>> res.statval
498.21750663761935
>>> res.dof
439
>>> res.parnames
('pl.gamma', 'pl.ampl', 'gline.fwhm', 'gline.pos', 'gline.ampl')
>>> res.parvals
(-0.20659543380329071, 0.00030398852609788524, 100.0, 4900.0, 0.001)
get_functions()[source] [edit on github]

Return the functions provided by Sherpa.

Returns

functions

Return type

list of str

See also

list_functions

Display the functions provided by Sherpa.

get_indep(id=None)[source] [edit on github]

Return the independent axes of a data set.

This function returns the coordinates of each point, or pixel, in the data set.

Parameters

id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

Returns

axis – The independent axis values. These are the values at which the model is evaluated during fitting.

Return type

tuple of arrays

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_dep

Return the dependent axis of a data set.

list_data_ids

List the identifiers for the loaded data sets.

Examples

For a one-dimensional data set, the X values are returned:

>>> load_arrays(1, [10, 15, 19], [4, 5, 9])
>>> get_indep()
(array([10, 15, 19]),)

For a 2D data set the X0 and X1 values are returned:

>>> x0 = [10, 15, 12, 19]
>>> x1 = [12, 14, 10, 17]
>>> y = [4, 5, 9, -2]
>>> load_arrays(2, x0, x1, y, Data2D)
>>> get_indep(2)
(array([10, 15, 12, 19]), array([12, 14, 10, 17]))
get_int_proj(par=None, id=None, otherids=None, recalc=False, fast=True, min=None, max=None, nloop=20, delv=None, fac=1, log=False, numcores=None)[source] [edit on github]

Return the interval-projection object.

This returns (and optionally calculates) the data used to display the int_proj plot. Note that if the the recalc parameter is False (the default value) then all other parameters are ignored and the results of the last int_proj call are returned.

Parameters
  • par – The parameter to plot. This argument is only used if recalc is set to True.

  • id (str or int, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (list of str or int, optional) – Other data sets to use in the calculation.

  • recalc (bool, optional) – The default value (False) means that the results from the last call to int_proj (or get_int_proj) are returned, ignoring all other parameter values. Otherwise, the statistic curve is re-calculated, but not plotted.

  • fast (bool, optional) – If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

  • min (number, optional) – The minimum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • max (number, optional) – The maximum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • nloop (int, optional) – The number of steps to use. This is used when delv is set to None.

  • delv (number, optional) – The step size for the parameter. Setting this over-rides the nloop parameter. The default is None.

  • fac (number, optional) – When min or max is not given, multiply the covariance of the parameter by this value to calculate the limit (which is then added or subtracted to the parameter value, as required).

  • log (bool, optional) – Should the step size be logarithmically spaced? The default (False) is to use a linear grid.

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

Returns

iproj – The fields of this object can be used to re-create the plot created by int_proj.

Return type

a sherpa.plot.IntervalProjection instance

See also

conf

Estimate parameter confidence intervals using the confidence method.

covar

Estimate the confidence intervals using the covariance method.

int_proj

Calculate and plot the fit statistic versus fit parameter value.

int_unc

Calculate and plot the fit statistic versus fit parameter value.

reg_proj

Plot the statistic value as two parameters are varied.

Examples

Return the results of the int_proj run:

>>> int_proj(src.xpos)
>>> iproj = get_int_proj()
>>> min(iproj.y)
119.55942437129544

Since the recalc parameter has not been changed to True, the following will return the results for the last call to int_proj, which may not have been for the src.ypos parameter:

>>> iproj = get_int_proj(src.ypos)

Create the data without creating a plot:

>>> iproj = get_int_proj(pl.gamma, recalc=True)

Specify the range and step size for the parameter, in this case varying linearly between 12 and 14 with 51 values:

>>> iproj = get_int_proj(src.r0, id="src", min=12, max=14,
...                      nloop=51, recalc=True)
get_int_unc(par=None, id=None, otherids=None, recalc=False, min=None, max=None, nloop=20, delv=None, fac=1, log=False, numcores=None)[source] [edit on github]

Return the interval-uncertainty object.

This returns (and optionally calculates) the data used to display the int_unc plot. Note that if the the recalc parameter is False (the default value) then all other parameters are ignored and the results of the last int_unc call are returned.

Parameters
  • par – The parameter to plot. This argument is only used if recalc is set to True.

  • id (str or int, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (list of str or int, optional) – Other data sets to use in the calculation.

  • recalc (bool, optional) – The default value (False) means that the results from the last call to int_proj (or get_int_proj) are returned, ignoring all other parameter values. Otherwise, the statistic curve is re-calculated, but not plotted.

  • min (number, optional) – The minimum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • max (number, optional) – The maximum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • nloop (int, optional) – The number of steps to use. This is used when delv is set to None.

  • delv (number, optional) – The step size for the parameter. Setting this over-rides the nloop parameter. The default is None.

  • fac (number, optional) – When min or max is not given, multiply the covariance of the parameter by this value to calculate the limit (which is then added or subtracted to the parameter value, as required).

  • log (bool, optional) – Should the step size be logarithmically spaced? The default (False) is to use a linear grid.

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

Returns

iunc – The fields of this object can be used to re-create the plot created by int_unc.

Return type

a sherpa.plot.IntervalUncertainty instance

See also

conf

Estimate parameter confidence intervals using the confidence method.

covar

Estimate the confidence intervals using the covariance method.

int_proj

Calculate and plot the fit statistic versus fit parameter value.

int_unc

Calculate and plot the fit statistic versus fit parameter value.

reg_proj

Plot the statistic value as two parameters are varied.

Examples

Return the results of the int_unc run:

>>> int_unc(src.xpos)
>>> iunc = get_int_unc()
>>> min(iunc.y)
119.55942437129544

Since the recalc parameter has not been changed to True, the following will return the results for the last call to int_unc, which may not have been for the src.ypos parameter:

>>> iunc = get_int_unc(src.ypos)

Create the data without creating a plot:

>>> iunc = get_int_unc(pl.gamma, recalc=True)

Specify the range and step size for the parameter, in this case varying linearly between 12 and 14 with 51 values:

>>> iunc = get_int_unc(src.r0, id="src", min=12, max=14,
...                    nloop=51, recalc=True)
get_iter_method_name()[source] [edit on github]

Return the name of the iterative fitting scheme.

Returns

name – The name of the iterative fitting scheme set by set_iter_method.

Return type

{‘none’, ‘sigmarej’}

See also

list_iter_methods

List the iterative fitting schemes.

set_iter_method

Set the iterative-fitting scheme used in the fit.

Examples

>>> print(get_iter_method_name())
get_iter_method_opt(optname=None)[source] [edit on github]

Return one or all options for the iterative-fitting scheme.

The options available for the iterative-fitting methods are described in set_iter_method_opt.

Parameters

optname (str, optional) – If not given, a dictionary of all the options are returned. When given, the individual value is returned.

Returns

value – The dictionary is empty when no iterative scheme is being used.

Return type

dictionary or value

Raises

sherpa.utils.err.ArgumentErr – If the optname argument is not recognized.

See also

get_iter_method_name

Return the name of the iterative fitting scheme.

set_iter_method_opt

Set an option for the iterative-fitting scheme.

set_iter_method

Set the iterative-fitting scheme used in the fit.

Examples

Display the settings of the current iterative-fitting method:

>>> print(get_iter_method_opt())

Switch to the sigmarej scheme and find out the current settings:

>>> set_iter_method('sigmarej')
>>> opts = get_iter_method_opt()

Return the ‘maxiters’ setting (if applicable):

>>> get_iter_method_opt('maxiters')
get_kernel_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_kernel.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_kernel (or get_kernel_contour) are returned, otherwise the data is re-generated.

Returns

psf_data

Return type

a sherpa.plot.PSFKernelContour instance

Raises

See also

get_psf_contour

Return the data used by contour_psf.

contour_kernel

Contour the kernel applied to the model of an image data set.

contour_psf

Contour the PSF applied to the model of an image data set.

Examples

Return the contour data for the kernel for the default data set:

>>> kplot = get_kernel_contour()
get_kernel_image(id=None)[source] [edit on github]

Return the data used by image_kernel.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

psf_data

Return type

a sherpa.image.PSFKernelImage instance

Raises

sherpa.utils.err.IdentifierErr – If a PSF model has not been created for the data set.

See also

get_psf_image

Return the data used by image_psf.

image_kernel

Display the 2D kernel for a data set in the image viewer.

image_psf

Display the 2D PSF model for a data set in the image viewer.

Examples

Return the image data for the kernel for the default data set:

>>> lplot = get_kernel_image()
>>> iplot.y.shape
(51, 51)
get_kernel_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_kernel.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_kernel (or get_kernel_plot) are returned, otherwise the data is re-generated.

Returns

kernel_plot

Return type

a sherpa.plot.PSFKernelPlot instance

Raises

sherpa.utils.err.IdentifierErr – If a PSF model has not been created for the data set.

See also

get_psf_plot

Return the data used by plot_psf.

plot_kernel

Plot the 1D kernel applied to a data set.

plot_psf

Plot the 1D PSF model applied to a data set.

Examples

Return the plot data and then create a plot with it:

>>> kplot = get_kernel_plot()
>>> kplot.plot()
get_method(name=None)[source] [edit on github]

Return an optimization method.

Parameters

name (str, optional) – If not given, the current method is returned, otherwise it should be one of the names returned by the list_methods function.

Returns

method – An object representing the optimization method.

Return type

object

Raises

sherpa.utils.err.ArgumentErr – If the name argument is not recognized.

See also

get_method_opt

Get the options for the current optimization method.

list_methods

List the supported optimization methods.

set_method

Change the optimization method.

set_method_opt

Change an option of the current optimization method.

Examples

The fields of the object returned by get_method can be used to view or change the method options.

>>> method = ui.get_method()
>>> print(method.name)
levmar
>>> print(method)
name    = levmar
ftol    = 1.19209289551e-07
xtol    = 1.19209289551e-07
gtol    = 1.19209289551e-07
maxfev  = None
epsfcn  = 1.19209289551e-07
factor  = 100.0
verbose = 0
>>> method.maxfev = 5000
get_method_name()[source] [edit on github]

Return the name of current Sherpa optimization method.

Returns

name – The name of the current optimization method, in lower case. This may not match the value sent to set_method because some methods can be set by multiple names.

Return type

str

See also

get_method

Return an optimization method.

get_method_opt

Get the options for the current optimization method.

Examples

>>> get_method_name()
'levmar'

The ‘neldermead’ method can also be referred to as ‘simplex’:

>>> set_method('simplex')
>>> get_method_name()
'neldermead'
get_method_opt(optname=None)[source] [edit on github]

Return one or all of the options for the current optimization method.

This is a helper function since the optimization options can also be read directly using the object returned by get_method.

Parameters

optname (str, optional) – If not given, a dictionary of all the options are returned. When given, the individual value is returned.

Returns

value

Return type

dictionary or value

Raises

sherpa.utils.err.ArgumentErr – If the optname argument is not recognized.

See also

get_method

Return an optimization method.

set_method

Change the optimization method.

set_method_opt

Change an option of the current optimization method.

Examples

>>> get_method_opt('maxfev') is None
True
>>> mopts = get_method_opt()
>>> mopts['maxfev'] is None
True
get_model(id=None)[source] [edit on github]

Return the model expression for a data set.

This returns the model expression for a data set, including any instrument response (e.g. PSF or ARF and RMF) whether created automatically or explicitly, with set_full_model.

Parameters

id (int or str, optional) – The data set containing the source expression. If not given then the default identifier is used, as returned by get_default_id.

Returns

This can contain multiple model components and any instrument response. Changing attributes of this model changes the model used by the data set.

Return type

instance

See also

delete_model

Delete the model expression from a data set.

get_model_pars

Return the names of the parameters of a model.

get_model_type

Describe a model expression.

get_source

Return the source model expression for a data set.

list_model_ids

List of all the data sets with a source expression.

sherpa.astro.ui.set_bkg_model

Set the background model expression for a data set.

set_model

Set the source model expression for a data set.

set_full_model

Define the convolved model expression for a data set.

show_model

Display the source model expression for a data set.

Examples

Return the model fitted to the default data set:

>>> mdl = get_model()
>>> len(mdl.pars)
5
get_model_autoassign_func()[source] [edit on github]

Return the method used to create model component identifiers.

Provides access to the function which is used by create_model_component and when creating model components directly to add an identifier in the current Python namespace.

Returns

The model function set by set_model_autoassign_func.

Return type

func

See also

create_model_component

Create a model component.

set_model

Set the source model expression for a data set.

set_model_autoassign_func

Set the method used to create model component identifiers.

get_model_component(name)[source] [edit on github]

Returns a model component given its name.

Parameters

name (str) – The name of the model component.

Returns

component – The model component object.

Return type

a sherpa.models.model.Model instance

Raises

sherpa.utils.err.IdentifierErr – If there is no model component with the given name.

See also

create_model_component

Create a model component.

get_model

Return the model expression for a data set.

get_source

Return the source model expression for a data set.

list_model_components

List the names of all the model components.

set_model

Set the source model expression for a data set.

Notes

The model instances are named as modeltype.username, and it is the username component that is used here to access the instance.

Examples

When a model component is created, a variable is created that contains the model instance. The instance can also be returned with get_model_component, which can then be queried or used to change the model settings:

>>> create_model_component('gauss1d', 'gline')
>>> gmodel = get_model_component('gline')
>>> gmodel.name
'gauss1d.gline'
>>> print([p.name for p in gmodel.pars])
['fwhm', 'pos', 'ampl']
>>> gmodel.fwhm.val = 12.2
>>> gmodel.fwhm.freeze()
get_model_component_image(id, model=None)[source] [edit on github]

Return the data used by image_model_component.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to display (the name, if a string).

Returns

cpt_img – The y attribute contains the component model values as a 2D NumPy array.

Return type

a sherpa.image.ComponentModelImage instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_source_component_image

Return the data used by image_source_component.

get_model_image

Return the data used by image_model.

image_model

Display the model for a data set in the image viewer.

image_model_component

Display a component of the model in the image viewer.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Examples

Return the gsrc component values for the default data set:

>>> minfo = get_model_component_image(gsrc)

Get the bgnd model pixel values for data set 2:

>>> minfo = get_model_component_image(2, bgnd)
get_model_component_plot(id, model=None, recalc=True)[source] [edit on github]

Return the data used to create the model-component plot.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to use (the name, if a string).

  • recalc (bool, optional) – If False then the results from the last call to plot_model_component (or get_model_component_plot) are returned, otherwise the data is re-generated.

Returns

An object representing the data used to create the plot by plot_model_component. The return value depends on the data set (e.g. 1D binned or un-binned).

Return type

instance

See also

get_model_plot

Return the data used to create the model plot.

plot_model

Plot the model for a data set.

plot_model_component

Plot a component of the model for a data set.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Examples

Return the plot data for the pl component used in the default data set:

>>> cplot = get_model_component_plot(pl)

Return the full source model (fplot) and then for the components gal * pl and gal * gline, for the data set ‘jet’:

>>> fmodel = xsphabs.gal * (powlaw1d.pl + gauss1d.gline)
>>> set_source('jet', fmodel)
>>> fit('jet')
>>> fplot = get_model_plot('jet')
>>> plot1 = get_model_component_plot('jet', pl*gal)
>>> plot2 = get_model_component_plot('jet', gline*gal)
get_model_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_model.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_model (or get_model_contour) are returned, otherwise the data is re-generated.

Returns

model_data – The y attribute contains the model values and the x0 and x1 arrays contain the corresponding coordinate values, as one-dimensional arrays.

Return type

a sherpa.plot.ModelContour instance

Raises

See also

get_model_image

Return the data used by image_model.

contour_model

Contour the values of the model, including any PSF.

image_model

Display the model for a data set in the image viewer.

Examples

Return the model pixel values for the default data set:

>>> minfo = get_model_contour()
get_model_contour_prefs()[source] [edit on github]

Return the preferences for contour_model.

Returns

prefs – Changing the values of this dictionary will change any new contour plots.

Return type

dict

See also

contour_model

Contour the values of the model, including any PSF.

Notes

The meaning of the fields depend on the chosen plot backend. A value of None (or not set) means to use the default value for that attribute, unless indicated otherwise.

alpha

The transparency value used to draw the contours, where 0 is fully transparent and 1 is fully opaque.

colors

The colors to draw the contours.

linewidths

What thickness of line to draw the contours.

xlog

Should the X axis be drawn with a logarithmic scale? The default is False.

ylog

Should the Y axis be drawn with a logarithmic scale? The default is False.

Examples

Change the contours for the model to be drawn in ‘orange’:

>>> prefs = get_model_contour_prefs()
>>> prefs['color'] = 'orange'
>>> contour_data()
>>> contour_model(overcontour=True)
get_model_image(id=None)[source] [edit on github]

Return the data used by image_model.

Evaluate the source expression for the image pixels - including any PSF convolution defined by set_psf - and return the results.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

src_img – The y attribute contains the source model values as a 2D NumPy array.

Return type

a sherpa.image.ModelImage instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_source_image

Return the data used by image_source.

contour_model

Contour the values of the model, including any PSF.

image_model

Display the model for a data set in the image viewer.

set_psf

Add a PSF model to a data set.

Examples

Calculate the residuals (data - model) for the default data set:

>>> minfo = get_model_image()
>>> dinfo = get_data_image()
>>> resid = dinfo.y - minfo.y
get_model_pars(model)[source] [edit on github]

Return the names of the parameters of a model.

Parameters

model (str or a sherpa.models.model.Model object) –

Returns

names – The names of the parameters in the model expression. These names do not include the name of the parent component.

Return type

list of str

See also

create_model_component

Create a model component.

get_model

Return the model expression for a data set.

get_model_type

Describe a model expression.

get_source

Return the source model expression for a data set.

Examples

>>> set_source(gauss2d.src + const2d.bgnd)
>>> get_model_pars(get_source())
['fwhm', 'xpos', 'ypos', 'ellip', 'theta', 'ampl', 'c0']
get_model_plot(id=None, recalc=True)[source] [edit on github]

Return the data used to create the model plot.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_model (or get_model_plot) are returned, otherwise the data is re-generated.

Returns

An object representing the data used to create the plot by plot_model. The return value depends on the data set (e.g. 1D binned or un-binned).

Return type

instance

See also

get_model_plot_prefs

Return the preferences for plot_model.

plot_model

Plot the model for a data set.

Examples

>>> mplot = get_model_plot()
>>> print(mplot)
get_model_plot_prefs(id=None)[source] [edit on github]

Return the preferences for plot_model.

The plot preferences may depend on the data set, so it is now an optional argument.

Changed in version 4.12.2: The id argument has been given.

Parameters

id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

Returns

prefs – Changing the values of this dictionary will change any new model plots. This dictionary will be empty if no plot backend is available.

Return type

dict

See also

plot_model

Plot the model for a data set.

set_xlinear

New plots will display a linear X axis.

set_xlog

New plots will display a logarithmically-scaled X axis.

set_ylinear

New plots will display a linear Y axis.

set_ylog

New plots will display a logarithmically-scaled Y axis.

Notes

The meaning of the fields depend on the chosen plot backend. A value of None means to use the default value for that attribute, unless indicated otherwise. These preferences are used by the following commands: plot_model, plot_ratio, plot_bkg_model, and the “fit” variants, such as plot_fit, plot_fit_resid, and plot_bkg_fit.

The preferences recognized by the matplotlib backend are the same as for get_data_plot_prefs.

Examples

After these commands, any model plot will use a green line to display the model:

>>> prefs = get_model_plot_prefs()
>>> prefs['color'] = 'green'
get_model_type(model)[source] [edit on github]

Describe a model expression.

Parameters

model (str or a sherpa.models.model.Model object) –

Returns

type – The name of the model expression.

Return type

str

See also

create_model_component

Create a model component.

get_model

Return the model expression for a data set.

get_model_pars

Return the names of the parameters of a model.

get_source

Return the source model expression for a data set.

Examples

>>> create_model_component("powlaw1d", "pl")
>>> get_model_type("pl")
'powlaw1d'

For expressions containing more than one component, the result is likely to be ‘binaryopmodel’

>>> get_model_type(const1d.norm * (polynom1d.poly + gauss1d.gline))
'binaryopmodel'

For sources with some form of an instrument model - such as a PSF convolution for an image or a PHA file with response information from the ARF and RMF - the response can depend on whether the expression contains this extra information or not:

>>> get_model_type(get_source('spec'))
'binaryopmodel'
>>> get_model_type(get_model('spec'))
'rspmodelpha'
get_num_par(id=None)[source] [edit on github]

Return the number of parameters in a model expression.

The get_num_par function returns the number of parameters, both frozen and thawed, in the model assigned to a data set.

Parameters

id (int or str, optional) – The data set containing the model expression. If not given then the default identifier is used, as returned by get_default_id.

Returns

npar – The number of parameters in the model expression. This sums up all the parameters of the components in the expression, and includes both frozen and thawed components.

Return type

int

Raises

sherpa.utils.err.IdentifierErr – If no model expression has been set for the data set (with set_model or set_source).

See also

get_num_par_frozen

Return the number of frozen parameters.

get_num_par_thawed

Return the number of thawed parameters.

set_model

Set the source model expression for a data set.

Examples

Return the total number of parameters for the default data set:

>>> print(get_num_par())

Find the number of parameters for the model associated with the data set called “jet”:

>>> njet = get_num_par('jet')
get_num_par_frozen(id=None)[source] [edit on github]

Return the number of frozen parameters in a model expression.

The get_num_par_frozen function returns the number of frozen parameters in the model assigned to a data set.

Parameters

id (int or str, optional) – The data set containing the model expression. If not given then the default identifier is used, as returned by get_default_id.

Returns

npar – The number of parameters in the model expression. This sums up all the frozen parameters of the components in the expression.

Return type

int

Raises

sherpa.utils.err.IdentifierErr – If no model expression has been set for the data set (with set_model or set_source).

See also

get_num_par

Return the number of parameters.

get_num_par_thawed

Return the number of thawed parameters.

set_model

Set the source model expression for a data set.

Examples

Return the number of frozen parameters for the default data set:

>>> print(get_num_par_frozen())

Find the number of frozen parameters for the model associated with the data set called “jet”:

>>> njet = get_num_par_frozen('jet')
get_num_par_thawed(id=None)[source] [edit on github]

Return the number of thawed parameters in a model expression.

The get_num_par_thawed function returns the number of thawed parameters in the model assigned to a data set.

Parameters

id (int or str, optional) – The data set containing the model expression. If not given then the default identifier is used, as returned by get_default_id.

Returns

npar – The number of parameters in the model expression. This sums up all the thawed parameters of the components in the expression.

Return type

int

Raises

sherpa.utils.err.IdentifierErr – If no model expression has been set for the data set (with set_model or set_source).

See also

get_num_par

Return the number of parameters.

get_num_par_frozen

Return the number of frozen parameters.

set_model

Set the source model expression for a data set.

Examples

Return the number of thawed parameters for the default data set:

>>> print(get_num_par_thawed())

Find the number of thawed parameters for the model associated with the data set called “jet”:

>>> njet = get_num_par_thawed('jet')
get_par(par)[source] [edit on github]

Return a parameter of a model component.

Parameters

par (str) – The name of the parameter, using the format “componentname.parametername”.

Returns

par – The parameter values - e.g. current value, limits, and whether it is frozen - can be changed using this object.

Return type

a sherpa.models.parameter.Parameter instance

Raises

sherpa.utils.err.ArgumentErr – If the par argument is invalid: the model component does not exist or the given model has no parameter with that name.

See also

set_par

Set the value, limits, or behavior of a model parameter.

Examples

Return the “c0” parameter of the “bgnd” model component and change it to be frozen:

>>> p = get_par('bgnd.c0')
>>> p.frozen = True
get_pdf_plot()[source] [edit on github]

Return the data used to plot the last PDF.

Returns

plot – An object containing the data used by the last call to plot_pdf. The fields will be None if the function has not been called.

Return type

a sherpa.plot.PDFPlot instance

See also

plot_pdf

Plot the probability density function of an array.

get_prior(par)[source] [edit on github]

Return the prior function for a parameter (MCMC).

The default behavior of the pyBLoCXS MCMC sampler (run by the get_draws function) is to use a flat prior for each parameter. The get_prior routine finds the current prior assigned to a parameter, and set_prior is used to change it.

Parameters

par (a sherpa.models.parameter.Parameter instance) – A parameter of a model instance.

Returns

The parameter prior set by a previous call to set_prior. This may be a function or model instance.

Return type

prior

Raises

ValueError – If a prior has not been set for the parameter.

See also

set_prior

Set the prior function to use with a parameter.

Examples

>>> prior = get_prior(bgnd.c0)
>>> print(prior)
get_proj()[source] [edit on github]

Return the confidence-interval estimation object.

Note

The conf function should be used instead of proj.

Returns

proj

Return type

object

See also

conf

Estimate parameter confidence intervals using the confidence method.

get_proj_opt

Return one or all of the options for the confidence interval method.

proj

Estimate confidence intervals for fit parameters.

set_proj_opt

Set an option of the proj estimation object.

Notes

The attributes of the object include:

eps

The precision of the calculated limits. The default is 0.01.

fast

If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

max_rstat

If the reduced chi square is larger than this value, do not use (only used with chi-square statistics). The default is 3.

maxfits

The maximum number of re-fits allowed (that is, when the remin filter is met). The default is 5.

maxiters

The maximum number of iterations allowed when bracketing limits, before stopping for that parameter. The default is 200.

numcores

The number of computer cores to use when evaluating results in parallel. This is only used if parallel is True. The default is to use all cores.

parallel

If there is more than one free parameter then the results can be evaluated in parallel, to reduce the time required. The default is True.

remin

The minimum difference in statistic value for a new fit location to be considered better than the current best fit (which starts out as the starting location of the fit at the time proj is called). The default is 0.01.

sigma

What is the error limit being calculated. The default is 1.

soft_limits

Should the search be restricted to the soft limits of the parameters (True), or can parameter values go out all the way to the hard limits if necessary (False). The default is False

tol

The tolerance for the fit. The default is 0.2.

Examples

>>> print(get_proj())
name        = projection
numcores    = 8
max_rstat   = 3
maxiters    = 200
soft_limits = False
eps         = 0.01
fast        = False
maxfits     = 5
remin       = 0.01
tol         = 0.2
sigma       = 1
parallel    = True
get_proj_opt(name=None)[source] [edit on github]

Return one or all of the options for the confidence interval method.

Note

The conf function should be used instead of proj.

This is a helper function since the options can also be read directly using the object returned by get_proj.

Parameters

name (str, optional) – If not given, a dictionary of all the options are returned. When given, the individual value is returned.

Returns

value

Return type

dictionary or value

Raises

sherpa.utils.err.ArgumentErr – If the name argument is not recognized.

See also

conf

Estimate parameter confidence intervals using the confidence method.

proj

Estimate confidence intervals for fit parameters.

get_proj

Return the confidence-interval estimation object.

set_proj_opt

Set an option of the proj estimation object.

Examples

>>> get_proj_opt('sigma')
1
>>> popts = get_proj_opt()
>>> popts['sigma']
1
get_proj_results()[source] [edit on github]

Return the results of the last proj run.

Note

The conf function should be used instead of proj.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no proj call has been made.

See also

conf

Estimate parameter confidence intervals using the confidence method.

proj

Estimate confidence intervals for fit parameters.

get_proj_opt

Return one or all of the options for the projection method.

set_proj_opt

Set an option of the proj estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘projection’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

Examples

>>> res = get_proj_results()
>>> print(res)
datasets    = ('src',)
methodname  = projection
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('bgnd.c0',)
parvals     = (9.1958148476800918,)
parmins     = (-2.0765029551804268,)
parmaxes    = (2.0765029551935186,)
nfits       = 0
get_projection_results() [edit on github]

Return the results of the last proj run.

Note

The conf function should be used instead of proj.

Returns

results

Return type

sherpa.fit.ErrorEstResults object

Raises

sherpa.utils.err.SessionErr – If no proj call has been made.

See also

conf

Estimate parameter confidence intervals using the confidence method.

proj

Estimate confidence intervals for fit parameters.

get_proj_opt

Return one or all of the options for the projection method.

set_proj_opt

Set an option of the proj estimation object.

Notes

The fields of the object include:

datasets

A tuple of the data sets used in the analysis.

methodname

This will be ‘projection’.

iterfitname

The name of the iterated-fit method used, if any.

fitname

The name of the optimization method used.

statname

The name of the fit statistic used.

sigma

The sigma value used to calculate the confidence intervals.

percent

The percentage of the signal contained within the confidence intervals (calculated from the sigma value assuming a normal distribution).

parnames

A tuple of the parameter names included in the analysis.

parvals

A tuple of the best-fit parameter values, in the same order as parnames.

parmins

A tuple of the lower error bounds, in the same order as parnames.

parmaxes

A tuple of the upper error bounds, in the same order as parnames.

nfits

The number of model evaluations.

Examples

>>> res = get_proj_results()
>>> print(res)
datasets    = ('src',)
methodname  = projection
iterfitname = none
fitname     = levmar
statname    = chi2gehrels
sigma       = 1
percent     = 68.2689492137
parnames    = ('bgnd.c0',)
parvals     = (9.1958148476800918,)
parmins     = (-2.0765029551804268,)
parmaxes    = (2.0765029551935186,)
nfits       = 0
get_psf(id=None)[source] [edit on github]

Return the PSF model defined for a data set.

Return the parameter settings for the PSF model assigned to the data set.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

psf

Return type

a sherpa.instrument.PSFModel instance

Raises

sherpa.utils.err.IdentifierErr – If no PSF model has been set for the data set.

See also

delete_psf

Delete the PSF model for a data set.

image_psf

Display the 2D PSF model for a data set in the image viewer.

list_psf_ids

List of all the data sets with a PSF.

load_psf

Create a PSF model.

plot_psf

Plot the 1D PSF model applied to a data set.

set_psf

Add a PSF model to a data set.

Examples

Change the size and center of the PSF for the default data set:

>>> psf = get_psf()
>>> psf.size = (21, 21)
>>> psf.center = (10, 10)
get_psf_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_psf.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_psf (or get_psf_contour) are returned, otherwise the data is re-generated.

Returns

psf_data

Return type

a sherpa.plot.PSFContour instance

Raises

See also

get_kernel_contour

Return the data used by contour_kernel.

contour_kernel

Contour the kernel applied to the model of an image data set.

contour_psf

Contour the PSF applied to the model of an image data set.

Examples

Return the contour data for the PSF for the default data set:

>>> cplot = get_psf_contour()
get_psf_image(id=None)[source] [edit on github]

Return the data used by image_psf.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

psf_data

Return type

a sherpa.image.PSFImage instance

Raises

sherpa.utils.err.IdentifierErr – If a PSF model has not been created for the data set.

See also

get_kernel_image

Return the data used by image_kernel.

image_kernel

Display the 2D kernel for a data set in the image viewer.

image_psf

Display the 2D PSF model for a data set in the image viewer.

Examples

Return the image data for the PSF for the default data set:

>>> iplot = get_psf_image()
>>> iplot.y.shape
(175, 200)
get_psf_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_psf.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_psf (or get_psf_plot) are returned, otherwise the data is re-generated.

Returns

psf_plot

Return type

a sherpa.plot.PSFPlot instance

Raises

sherpa.utils.err.IdentifierErr – If a PSF model has not been created for the data set.

See also

get_kernel_plot

Return the data used by plot_kernel.

plot_kernel

Plot the 1D kernel applied to a data set.

plot_psf

Plot the 1D PSF model applied to a data set.

Examples

Return the plot data and then create a plot with it:

>>> pplot = get_psf_plot()
>>> pplot.plot()
get_pvalue_plot(null_model=None, alt_model=None, conv_model=None, id=1, otherids=(), num=500, bins=25, numcores=None, recalc=False)[source] [edit on github]

Return the data used by plot_pvalue.

Access the data arrays and preferences defining the histogram plot produced by the plot_pvalue function, a histogram of the likelihood ratios comparing fits of the null model to fits of the alternative model using faked data with Poisson noise. Data returned includes the likelihood ratio computed using the observed data, and the p-value, used to reject or accept the null model.

Parameters
  • null_model – The model expression for the null hypothesis.

  • alt_model – The model expression for the alternative hypothesis.

  • conv_model (optional) – An expression used to modify the model so that it can be compared to the data (e.g. a PSF or PHA response).

  • id (int or str, optional) – The data set that provides the data. The default is 1.

  • otherids (sequence of int or str, optional) – Other data sets to use in the calculation.

  • num (int, optional) – The number of simulations to run. The default is 500.

  • bins (int, optional) – The number of bins to use to create the histogram. The default is 25.

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

  • recalc (bool, optional) – The default value (False) means that the results from the last call to plot_pvalue or get_pvalue_plot are returned. If True, the values are re-calculated.

Returns

plot

Return type

a sherpa.plot.LRHistogram instance

See also

get_pvalue_results

Return the data calculated by the last plot_pvalue call.

plot_pvalue

Compute and plot a histogram of likelihood ratios by simulating data.

Examples

Return the values from the last call to plot_pvalue:

>>> pvals = get_pvalue_plot()
>>> pvals.ppp
0.472

Run 500 simulations for the two models and print the results:

>>> pvals = get_pvalue_plot(mdl1, mdl2, recalc=True, num=500)
>>> print(pvals)
get_pvalue_results()[source] [edit on github]

Return the data calculated by the last plot_pvalue call.

The get_pvalue_results function returns the likelihood ratio test results computed by the plot_pvalue command, which compares fits of the null model to fits of the alternative model using faked data with Poisson noise. The likelihood ratio based on the observed data is returned, along with the p-value, used to reject or accept the null model.

Returns

plot – If plot_pvalue or get_pvalue_plot have been called then the return value is a sherpa.sim.simulate.LikelihoodRatioResults instance, otherwise None is returned.

Return type

None or a sherpa.sim.simulate.LikelihoodRatioResults instance

See also

plot_value

Compute and plot a histogram of likelihood ratios by simulating data.

get_pvalue_plot

Return the data used by plot_pvalue.

Notes

The fields of the returned (LikelihoodRatioResults) object are:

ratios

The calculated likelihood ratio for each iteration.

stats

The calculated fit statistics for each iteration, stored as the null model and then the alt model in a nsim by 2 array.

samples

The parameter samples array for each simulation, stored in a nsim by npar array.

lr

The likelihood ratio of the observed data for the null and alternate models.

ppp

The p value of the observed data for the null and alternate models.

null

The fit statistic of the null model on the observed data.

alt

The fit statistic of the alternate model on the observed data.

Examples

Return the results of the last pvalue analysis and display the results - first using the format method, which provides a summary of the data, and then a look at the individual fields in the returned object. The last call displays the contents of one of the fields (ppp).

>>> res = get_pvalue_results()
>>> print(res.format())
>>> print(res)
>>> print(res.ppp)
get_ratio_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_ratio.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_ratio (or get_ratio_contour) are returned, otherwise the data is re-generated.

Returns

ratio_data – The y attribute contains the ratio values and the x0 and x1 arrays contain the corresponding coordinate values, as one-dimensional arrays.

Return type

a sherpa.plot.RatioContour instance

Raises

See also

get_ratio_image

Return the data used by image_ratio.

get_resid_contour

Return the data used by contour_resid.

contour_ratio

Contour the ratio of data to model.

image_ratio

Display the ratio (data/model) for a data set in the image viewer.

Examples

Return the ratio data for the default data set:

>>> rinfo = get_ratio_contour()
get_ratio_image(id=None)[source] [edit on github]

Return the data used by image_ratio.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

ratio_img – The y attribute contains the ratio values as a 2D NumPy array.

Return type

a sherpa.image.RatioImage instance

Raises

See also

get_resid_image

Return the data used by image_resid.

contour_ratio

Contour the ratio of data to model.

image_ratio

Display the ratio (data/model) for a data set in the image viewer.

Examples

Return the ratio data for the default data set:

>>> rinfo = get_ratio_image()
get_ratio_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_ratio.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_ratio (or get_ratio_plot) are returned, otherwise the data is re-generated.

Returns

ratio_data

Return type

a sherpa.plot.RatioPlot instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_chisqr_plot

Return the data used by plot_chisqr.

get_delchi_plot

Return the data used by plot_delchi.

get_resid_plot

Return the data used by plot_resid.

plot_ratio

Plot the ratio of data to model for a data set.

Examples

Return the ratio of the data to the model for the default data set:

>>> rplot = get_ratio_plot()
>>> np.min(rplot.y)
0.6320905073750186
>>> np.max(rplot.y)
1.5170172177000447

Display the contents of the ratio plot for data set 2:

>>> print(get_ratio_plot(2))

Overplot the ratio plot from the ‘core’ data set on the ‘jet’ data set:

>>> r1 = get_ratio_plot('jet')
>>> r2 = get_ratio_plot('core')
>>> r1.plot()
>>> r2.overplot()
get_reg_proj(par0=None, par1=None, id=None, otherids=None, recalc=False, fast=True, min=None, max=None, nloop=(10, 10), delv=None, fac=4, log=(False, False), sigma=(1, 2, 3), levels=None, numcores=None)[source] [edit on github]

Return the region-projection object.

This returns (and optionally calculates) the data used to display the reg_proj contour plot. Note that if the the recalc parameter is False (the default value) then all other parameters are ignored and the results of the last reg_proj call are returned.

Parameters
  • par0 – The parameters to plot on the X and Y axes, respectively. These arguments are only used if recalc is set to True.

  • par1 – The parameters to plot on the X and Y axes, respectively. These arguments are only used if recalc is set to True.

  • id (str or int, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (list of str or int, optional) – Other data sets to use in the calculation.

  • recalc (bool, optional) – The default value (False) means that the results from the last call to reg_proj (or get_reg_proj) are returned, ignoring all other parameter values. Otherwise, the statistic curve is re-calculated, but not plotted.

  • fast (bool, optional) – If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

  • min (pair of numbers, optional) – The minimum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • max (pair of number, optional) – The maximum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • nloop (pair of int, optional) – The number of steps to use. This is used when delv is set to None.

  • delv (pair of number, optional) – The step size for the parameter. Setting this over-rides the nloop parameter. The default is None.

  • fac (number, optional) – When min or max is not given, multiply the covariance of the parameter by this value to calculate the limit (which is then added or subtracted to the parameter value, as required).

  • log (pair of bool, optional) – Should the step size be logarithmically spaced? The default (False) is to use a linear grid.

  • sigma (sequence of number, optional) – The levels at which to draw the contours. The units are the change in significance relative to the starting value, in units of sigma.

  • levels (sequence of number, optional) – The numeric values at which to draw the contours. This over-rides the sigma parameter, if set (the default is None).

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

Returns

rproj – The fields of this object can be used to re-create the plot created by reg_proj.

Return type

a sherpa.plot.RegionProjection instance

See also

conf

Estimate patameter confidence intervals using the confidence method.

covar

Estimate the confidence intervals using the covariance method.

int_proj

Calculate and plot the fit statistic versus fit parameter value.

int_unc

Calculate and plot the fit statistic versus fit parameter value.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

Examples

Return the results for the reg_proj run for the xpos and ypos parameters of the src component, for the default data set:

>>> reg_proj(src.xpos, src.ypos)
>>> rproj = get_reg_proj()

Since the recalc parameter has not been changed to True, the following will return the results for the last call to reg_proj, which may not have been for the r0 and alpha parameters:

>>> rprog = get_reg_proj(src.r0, src.alpha)

Create the data without creating a plot:

>>> rproj = get_reg_proj(pl.gamma, gal.nh, recalc=True)

Specify the range and step size for both the parameters, in this case pl.gamma should vary between 0.5 and 2.5, with gal.nh between 0.01 and 1, both with 51 values and the nH range done over a log scale:

>>> rproj = get_reg_proj(pl.gamma, gal.nh, id="src",
...                      min=(0.5, 0.01), max=(2.5, 1),
...                      nloop=(51, 51), log=(False, True),
...                      recalc=True)
get_reg_unc(par0=None, par1=None, id=None, otherids=None, recalc=False, min=None, max=None, nloop=(10, 10), delv=None, fac=4, log=(False, False), sigma=(1, 2, 3), levels=None, numcores=None)[source] [edit on github]

Return the region-uncertainty object.

This returns (and optionally calculates) the data used to display the reg_unc contour plot. Note that if the the recalc parameter is False (the default value) then all other parameters are ignored and the results of the last reg_unc call are returned.

Parameters
  • par0 – The parameters to plot on the X and Y axes, respectively. These arguments are only used if recalc is set to True.

  • par1 – The parameters to plot on the X and Y axes, respectively. These arguments are only used if recalc is set to True.

  • id (str or int, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (list of str or int, optional) – Other data sets to use in the calculation.

  • recalc (bool, optional) – The default value (False) means that the results from the last call to reg_unc (or get_reg_unc) are returned, ignoring all other parameter values. Otherwise, the statistic curve is re-calculated, but not plotted.

  • fast (bool, optional) – If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

  • min (pair of numbers, optional) – The minimum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • max (pair of number, optional) – The maximum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • nloop (pair of int, optional) – The number of steps to use. This is used when delv is set to None.

  • delv (pair of number, optional) – The step size for the parameter. Setting this over-rides the nloop parameter. The default is None.

  • fac (number, optional) – When min or max is not given, multiply the covariance of the parameter by this value to calculate the limit (which is then added or subtracted to the parameter value, as required).

  • log (pair of bool, optional) – Should the step size be logarithmically spaced? The default (False) is to use a linear grid.

  • sigma (sequence of number, optional) – The levels at which to draw the contours. The units are the change in significance relative to the starting value, in units of sigma.

  • levels (sequence of number, optional) – The numeric values at which to draw the contours. This over-rides the sigma parameter, if set (the default is None).

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

Returns

rproj – The fields of this object can be used to re-create the plot created by reg_unc.

Return type

a sherpa.plot.RegionUncertainty instance

See also

conf

Estimate patameter confidence intervals using the confidence method.

covar

Estimate the confidence intervals using the covariance method.

int_proj

Calculate and plot the fit statistic versus fit parameter value.

int_unc

Calculate and plot the fit statistic versus fit parameter value.

reg_proj

Plot the statistic value as two parameters are varied.

reg_unc

Plot the statistic value as two parameters are varied.

Examples

Return the results for the reg_unc run for the xpos and ypos parameters of the src component, for the default data set:

>>> reg_unc(src.xpos, src.ypos)
>>> runc = get_reg_unc()

Since the recalc parameter has not been changed to True, the following will return the results for the last call to reg_unc, which may not have been for the r0 and alpha parameters:

>>> runc = get_reg_unc(src.r0, src.alpha)

Create the data without creating a plot:

>>> runc = get_reg_unc(pl.gamma, gal.nh, recalc=True)

Specify the range and step size for both the parameters, in this case pl.gamma should vary between 0.5 and 2.5, with gal.nh between 0.01 and 1, both with 51 values and the nH range done over a log scale:

>>> runc = get_reg_unc(pl.gamma, gal.nh, id="src",
...                    min=(0.5, 0.01), max=(2.5, 1),
...                    nloop=(51, 51), log=(False, True),
...                    recalc=True)
get_resid_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_resid.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_resid (or get_resid_contour) are returned, otherwise the data is re-generated.

Returns

resid_data – The y attribute contains the residual values and the x0 and x1 arrays contain the corresponding coordinate values, as one-dimensional arrays.

Return type

a sherpa.plot.ResidContour instance

Raises

See also

get_ratio_contour

Return the data used by contour_ratio.

get_resid_image

Return the data used by image_resid.

contour_resid

Contour the residuals of the fit.

image_resid

Display the residuals (data - model) for a data set in the image viewer.

Examples

Return the residual data for the default data set:

>>> rinfo = get_resid_contour()
get_resid_image(id=None)[source] [edit on github]

Return the data used by image_resid.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

resid_img – The y attribute contains the residual values as a 2D NumPy array.

Return type

a sherpa.image.ResidImage instance

Raises

See also

get_ratio_image

Return the data used by image_ratio.

contour_resid

Contour the residuals of the fit.

image_resid

Display the residuals (data - model) for a data set in the image viewer.

Examples

Return the residual data for the default data set:

>>> rinfo = get_resid_image()
get_resid_plot(id=None, recalc=True)[source] [edit on github]

Return the data used by plot_resid.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_resid (or get_resid_plot) are returned, otherwise the data is re-generated.

Returns

resid_data

Return type

a sherpa.plot.ResidPlot instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_chisqr_plot

Return the data used by plot_chisqr.

get_delchi_plot

Return the data used by plot_delchi.

get_ratio_plot

Return the data used by plot_ratio.

plot_resid

Plot the residuals (data - model) for a data set.

Examples

Return the residual data for the default data set:

>>> rplot = get_resid_plot()
>>> np.min(rplot.y)
-2.9102595936209896
>>> np.max(rplot.y)
4.0897404063790104

Display the contents of the residuals plot for data set 2:

>>> print(get_resid_plot(2))

Overplot the residuals plot from the ‘core’ data set on the ‘jet’ data set:

>>> r1 = get_resid_plot('jet')
>>> r2 = get_resid_plot('core')
>>> r1.plot()
>>> r2.overplot()
get_sampler()[source] [edit on github]

Return the current MCMC sampler options.

Returns the options for the current pyBLoCXS MCMC sampling method (jumping rules).

Returns

options – A copy of the options for the chosen sampler. Use set_sampler_opt to change these values. The fields depend on the current sampler.

Return type

dict

See also

get_sampler_name

Return the name of the current MCMC sampler.

get_sampler_opt

Return an option of the current MCMC sampler.

set_sampler

Set the MCMC sampler.

set_sampler_opt

Set an option for the current MCMC sampler.

Examples

>>> print(get_sampler())
get_sampler_name()[source] [edit on github]

Return the name of the current MCMC sampler.

Returns

name

Return type

str

See also

get_sampler

Return the current MCMC sampler options.

set_sampler

Set the MCMC sampler.

Examples

>>> get_sampler_name()
'MetropolisMH'
get_sampler_opt(opt)[source] [edit on github]

Return an option of the current MCMC sampler.

Returns

opt – The name of the option. The fields depend on the current sampler.

Return type

str

See also

get_sampler

Return the current MCMC sampler options.

set_sampler_opt

Set an option for the current MCMC sampler.

Examples

>>> get_sampler_opt('log')
False
get_scatter_plot()[source] [edit on github]

Return the data used to plot the last scatter plot.

Returns

plot – An object containing the data used by the last call to plot_scatter. The fields will be None if the function has not been called.

Return type

a sherpa.plot.ScatterPlot instance

See also

plot_scatter

Create a scatter plot.

get_source(id=None)[source] [edit on github]

Return the source model expression for a data set.

This returns the model expression created by set_model or set_source. It does not include any instrument response.

Parameters

id (int or str, optional) – The data set containing the source expression. If not given then the default identifier is used, as returned by get_default_id.

Returns

model – This can contain multiple model components. Changing attributes of this model changes the model used by the data set.

Return type

a sherpa.models.Model object

See also

delete_model

Delete the model expression from a data set.

get_model

Return the model expression for a data set.

get_model_pars

Return the names of the parameters of a model.

get_model_type

Describe a model expression.

list_model_ids

List of all the data sets with a source expression.

sherpa.astro.ui.set_bkg_model

Set the background model expression for a data set.

set_model

Set the source model expression for a data set.

set_full_model

Define the convolved model expression for a data set.

show_model

Display the source model expression for a data set.

Examples

Return the source expression for the default data set, display it, and then find the number of parameters in it:

>>> src = get_source()
>>> print(src)
>>> len(src.pars)
5

Set the source expression for data set ‘obs2’ to be equal to the model of data set ‘obs1’ multiplied by a scalar value:

>>> set_source('obs2', const1d.norm * get_source('obs1'))
get_source_component_image(id, model=None)[source] [edit on github]

Return the data used by image_source_component.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to display (the name, if a string).

Returns

cpt_img – The y attribute contains the component model values as a 2D NumPy array.

Return type

a sherpa.image.ComponentSourceImage instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_model_component_image

Return the data used by image_model_component.

get_source_image

Return the data used by image_source.

image_source

Display the source expression for a data set in the image viewer.

image_source_component

Display a component of the source expression in the image viewer.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Examples

Return the gsrc component values for the default data set:

>>> sinfo = get_source_component_image(gsrc)

Get the ‘bgnd’ model pixel values for data set 2:

>>> sinfo = get_source_component_image(2, bgnd)
get_source_component_plot(id, model=None, recalc=True)[source] [edit on github]

Return the data used by plot_source_component.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to use (the name, if a string).

  • recalc (bool, optional) – If False then the results from the last call to plot_source_component (or get_source_component_plot) are returned, otherwise the data is re-generated.

Returns

An object representing the data used to create the plot by plot_source_component. The return value depends on the data set (e.g. 1D binned or un-binned).

Return type

instance

See also

get_source_plot

Return the data used to create the source plot.

plot_source

Plot the source expression for a data set.

plot_source_component

Plot a component of the source expression for a data set.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Examples

Return the plot data for the pl component used in the default data set:

>>> cplot = get_source_component_plot(pl)

Return the full source model (fplot) and then for the components gal * pl and gal * gline, for the data set ‘jet’:

>>> fmodel = xsphabs.gal * (powlaw1d.pl + gauss1d.gline)
>>> set_source('jet', fmodel)
>>> fit('jet')
>>> fplot = get_source('jet')
>>> plot1 = get_source_component_plot('jet', pl*gal)
>>> plot2 = get_source_component_plot('jet', gline*gal)
get_source_contour(id=None, recalc=True)[source] [edit on github]

Return the data used by contour_source.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to contour_source (or get_source_contour) are returned, otherwise the data is re-generated.

Returns

source_data – The y attribute contains the model values and the x0 and x1 arrays contain the corresponding coordinate values, as one-dimensional arrays.

Return type

a sherpa.plot.SourceContour instance

Raises

See also

get_source_image

Return the data used by image_source.

contour_source

Contour the values of the model, without any PSF.

image_source

Display the source expression for a data set in the image viewer.

Examples

Return the source model pixel values for the default data set:

>>> sinfo = get_source_contour()
get_source_image(id=None)[source] [edit on github]

Return the data used by image_source.

Evaluate the source expression for the image pixels - without any PSF convolution - and return the results.

Parameters

id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

Returns

src_img – The y attribute contains the source model values as a 2D NumPy array.

Return type

a sherpa.image.SourceImage instance

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_model_image

Return the data used by image_model.

contour_source

Contour the values of the model, without any PSF.

image_source

Display the source expression for a data set in the image viewer.

Examples

Return the model data for the default data set:

>>> sinfo = get_source_image()
>>> sinfo.y.shape
(150, 175)
get_source_plot(id=None, recalc=True)[source] [edit on github]

Return the data used to create the source plot.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • recalc (bool, optional) – If False then the results from the last call to plot_source (or get_source_plot) are returned, otherwise the data is re-generated.

Returns

An object representing the data used to create the plot by plot_source. The return value depends on the data set (e.g. 1D binned or un-binned).

Return type

instance

See also

get_model_plot

Return the data used to create the model plot.

plot_model

Plot the model for a data set.

plot_source

Plot the source expression for a data set.

Examples

Retrieve the source plot information for the default data set and then display it:

>>> splot = get_source_plot()
>>> print(splot)

Return the plot data for data set 2, and then use it to create a plot:

>>> s2 = get_source_plot(2)
>>> s2.plot()

Display the two source plots for the ‘jet’ and ‘core’ datasets on the same plot:

>>> splot1 = get_source_plot(id='jet')
>>> splot2 = get_source_plot(id='core')
>>> splot1.plot()
>>> splot2.overplot()
get_split_plot()[source] [edit on github]

Return the plot attributes for displays with multiple plots.

Returns

splot

Return type

a sherpa.plot.SplitPlot instance

get_stat(name=None)[source] [edit on github]

Return the fit statisic.

Parameters

name (str, optional) – If not given, the current fit statistic is returned, otherwise it should be one of the names returned by the list_stats function.

Returns

stat – An object representing the fit statistic.

Return type

object

Raises

sherpa.utils.err.ArgumentErr – If the name argument is not recognized.

See also

get_stat_name

Return the name of the current fit statistic.

list_stats

List the fit statistics.

set_stat

Change the fit statistic.

Examples

Return the currently-selected statistic, display its name, and read the help documentation for it:

>>> stat = get_stat()
>>> stat.name
'chi2gehrels'
>>> help(stat)

Read the help for the “wstat” statistic:

>>> help(get_stat('wstat'))
get_stat_info()[source] [edit on github]

Return the statistic values for the current models.

Calculate the statistic value for each data set, and the combined fit, using the current set of models, parameters, and ranges.

Returns

stats – The values for each data set. If there are multiple model expressions then the last element will be the value for the combined data sets.

Return type

array of sherpa.fit.StatInfoResults

See also

calc_stat

Calculate the fit statistic for a data set.

calc_stat_info

Display the statistic values for the current models.

get_fit_results

Return the results of the last fit.

list_data_ids

List the identifiers for the loaded data sets.

list_model_ids

List of all the data sets with a source expression.

Notes

If a fit to a particular data set has not been made, or values - such as parameter settings, the noticed data range, or choice of statistic - have been changed since the last fit, then the results for that data set may not be meaningful and will therefore bias the results for the simultaneous results.

The return value of get_stat_info differs to get_fit_results since it includes values for each data set, individually, rather than just the combined results.

The fields of the object include:

name

The name of the data set, or sets, as a string.

ids

A sequence of the data set ids (it may be a tuple or array) included in the results.

bkg_ids

A sequence of the background data set ids (it may be a tuple or array) included in the results, if any.

statname

The name of the statistic function (as used in set_stat).

statval

The statistic value.

numpoints

The number of bins used in the fits.

dof

The number of degrees of freedom in the fit (the number of bins minus the number of free parameters).

qval

The Q-value (probability) that one would observe the reduced statistic value, or a larger value, if the assumed model is true and the current model parameters are the true parameter values. This will be None if the value can not be calculated with the current statistic (e.g. the Cash statistic).

rstat

The reduced statistic value (the statval field divided by dof). This is not calculated for all statistics.

Examples

>>> res = get_stat_info()
>>> res[0].statval
498.21750663761935
>>> res[0].dof
439
get_stat_name()[source] [edit on github]

Return the name of the current fit statistic.

Returns

name – The name of the current fit statistic method, in lower case.

Return type

str

See also

get_stat

Return a fit statistic.

set_stat

Set the fit statistic.

Examples

>>> get_stat_name()
'chi2gehrels'
>>> set_stat('cash')
>>> get_stat_name()
'cash'
get_staterror(id=None, filter=False)[source] [edit on github]

Return the statistical error on the dependent axis of a data set.

The function returns the statistical errors on the values (dependenent axis) of a data set. These may have been set explicitly - either when the data set was created or with a call to set_staterror - or as defined by the chosen fit statistic (such as “chi2gehrels”).

Parameters
  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • filter (bool, optional) – Should the filter attached to the data set be applied to the return value or not. The default is False.

Returns

staterrors – The statistical error for each data point. This may be estimated from the data (e.g. with the chi2gehrels statistic) or have been set explicitly (set_staterror). The size of this array depends on the filter argument.

Return type

array

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_error

Return the errors on the dependent axis of a data set.

get_indep

Return the independent axis of a data set.

get_syserror

Return the systematic errors on the dependent axis of a data set.

list_data_ids

List the identifiers for the loaded data sets.

set_staterror

Set the statistical errors on the dependent axis of a data set.

Notes

The default behavior is to not apply any filter defined on the independent axes to the results, so that the return value is for all points (or bins) in the data set. Set the filter argument to True to apply this filter.

Examples

If not explicitly given, the statistical errors on a data set may be calculated from the data values (the independent axis), depending on the chosen statistic:

>>> load_arrays(1, [10, 15, 19], [4, 5, 9])
>>> set_stat('chi2datavar')
>>> get_staterror()
array([ 2.        ,  2.23606798,  3.        ])
>>> set_stat('chi2gehrels')
>>> get_staterror()
array([ 3.17944947,  3.39791576,  4.122499  ])

If the statistical errors are set - either when the data set is created or with a call to set_staterror - then these values will be used, no matter the statistic:

>>> load_arrays(1, [10, 15, 19], [4, 5, 9], [2, 3, 5])
>>> set_stat('chi2datavar')
>>> get_staterror()
array([2, 3, 5])
>>> set_stat('chi2gehrels')
>>> get_staterror()
array([2, 3, 5])
get_syserror(id=None, filter=False)[source] [edit on github]

Return the systematic error on the dependent axis of a data set.

The function returns the systematic errors on the values (dependenent axis) of a data set. It is an error if called on a data set with no systematic errors (which are set with set_syserror).

Parameters
  • id (int or str, optional) – The identifier for the data set to use. If not given then the default identifier is used, as returned by get_default_id.

  • filter (bool, optional) – Should the filter attached to the data set be applied to the return value or not. The default is False.

Returns

syserrors – The systematic error for each data point. The size of this array depends on the filter argument.

Return type

array

Raises

See also

get_error

Return the errors on the dependent axis of a data set.

get_indep

Return the independent axis of a data set.

get_staterror

Return the statistical errors on the dependent axis of a data set.

list_data_ids

List the identifiers for the loaded data sets.

set_syserror

Set the systematic errors on the dependent axis of a data set.

Notes

The default behavior is to not apply any filter defined on the independent axes to the results, so that the return value is for all points (or bins) in the data set. Set the filter argument to True to apply this filter.

Examples

Return the systematic error for the default data set:

>>> yerr = get_syserror()

Return an array that has been filtered to match the data:

>>> yerr = get_syserror(filter=True)

Return the filtered errors for data set “core”:

>>> yerr = get_syserror("core", filter=True)
get_trace_plot()[source] [edit on github]

Return the data used to plot the last trace.

Returns

plot – An object containing the data used by the last call to plot_trace. The fields will be None if the function has not been called.

Return type

a sherpa.plot.TracePlot instance

See also

plot_trace

Create a trace plot of row number versus value.

guess(id=None, model=None, limits=True, values=True)[source] [edit on github]

Estimate the parameter values and ranges given the loaded data.

The guess function can change the parameter values and limits to match the loaded data. This is generally limited to changing the amplitude and position parameters (sometimes just the values and sometimes just the limits). The parameters that are changed depend on the type of model.

Parameters
  • id (int or str, optional) – The data set that provides the data. If not given then the default identifier is used, as returned by get_default_id.

  • model – Change the parameters of this model component. If None, then the source expression is assumed to consist of a single component, and that component is used.

  • limits (bool) – Should the parameter limits be changed? The default is True.

  • values (bool) – Should the parameter values be changed? The default is True.

See also

get_default_id

Return the default data set identifier.

reset

Reset the model parameters to their default settings.

set_par

Set the value, limits, or behavior of a model parameter.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

The guess function can reduce the time required to fit a data set by moving the parameters closer to a realistic solution. It can also be useful because it can set bounds on the parameter values based on the data: for instance, many two-dimensional models will limit their xpos and ypos values to lie within the data area. This can be done manually, but guess simplifies this, at least for those parameters that are supported. Instrument models - such as an ARF and RMF - should be set up before calling guess.

Examples

Since the source expression contains only one component, guess can be called with no arguments:

>>> set_source(polynom1d.poly)
>>> guess()

In this case, guess is called on each component separately.

>>> set_source(gauss1d.line + powlaw1d.cont)
>>> guess(line)
>>> guess(cont)

In this example, the values of the src model component are guessed from the “src” data set, whereas the bgnd component is guessed from the “bgnd” data set.

>>> set_source("src", gauss2d.src + const2d.bgnd)
>>> set_source("bgnd", bgnd)
>>> guess("src", src)
>>> guess("bgnd", bgnd)

Set the source model for the default dataset. Guess is run to determine the values of the model component “p1” and the limits of the model component “g1”:

>>> set_source(powlaw1d.p1 + gauss1d.g1)
>>> guess(p1, limits=False)
>>> guess(g1, values=False)
ignore(lo=None, hi=None, **kwargs)[source] [edit on github]

Exclude data from the fit.

Select one or more ranges of data to exclude by filtering on the independent axis value. The filter is applied to all data sets.

Changed in version 4.14.0: Integrated data sets - so Data1DInt and DataPHA when using energy or wavelengths - now ensure that the hi argument is exclusive and better handling of the lo argument when it matches a bin edge. This can result in the same filter selecting a smaller number of bins than in earlier versions of Sherpa.

Parameters
  • lo (number or str, optional) – The lower bound of the filter (when a number) or a string expression listing ranges in the form a:b, with multiple ranges allowed, where the ranges are separated by a ,. The term :b means exclude everything up to b (an exclusive limit for integrated datasets), and a: means exclude everything that is higher than, or equal to, a.

  • hi (number, optional) – The upper bound of the filter when lo is not a string.

  • bkg_id (int or str, optional) – The filter will be applied to the associated background component of the data set if bkg_id is set. Only PHA data sets support this option; if not given, then the filter is applied to all background components as well as the source data.

See also

ignore_id

Exclude data from the fit for a data set.

sherpa.astro.ui.ignore2d

Exclude a spatial region from an image.

notice

Include data in the fit.

show_filter

Show any filters applied to a data set.

Notes

The order of ignore and notice calls is important, and the results are a union, rather than intersection, of the combination.

For binned data sets, the bin is excluded if the ignored range falls anywhere within the bin.

The units used depend on the analysis setting of the data set, if appropriate.

To filter a 2D data set by a shape use ignore2d.

Examples

Ignore all data points with an X value (the independent axis) between 12 and 18. For this one-dimensional data set, this means that the second bin is ignored:

>>> load_arrays(1, [10, 15, 20, 30], [5, 10, 7, 13])
>>> ignore(12, 18)
>>> get_dep(filter=True)
array([ 5,  7, 13])

Filtering X values that are 25 or larger means that the last point is also ignored:

>>> ignore(25, None)
>>> get_dep(filter=True)
array([ 5,  7])

The notice call removes the previous filter, and then a multi-range filter is applied to exclude values between 8 and 12 and 18 and 22:

>>> notice()
>>> ignore("8:12,18:22")
>>> get_dep(filter=True)
array([10, 13])
ignore_id(ids, lo=None, hi=None, **kwargs)[source] [edit on github]

Exclude data from the fit for a data set.

Select one or more ranges of data to exclude by filtering on the independent axis value. The filter is applied to the given data set, or sets.

Changed in version 4.14.0: Integrated data sets - so Data1DInt and DataPHA when using energy or wavelengths - now ensure that the hi argument is exclusive and better handling of the lo argument when it matches a bin edge. This can result in the same filter selecting a smaller number of bins than in earlier versions of Sherpa.

Parameters
  • ids (int or str, or array of int or str) – The data set, or sets, to use.

  • lo (number or str, optional) – The lower bound of the filter (when a number) or a string expression listing ranges in the form a:b, with multiple ranges allowed, where the ranges are separated by a ,. The term :b means exclude everything up to b (an exclusive limit for integrated datasets), and a: means exclude everything that is higher than, or equal to, a.

  • hi (number, optional) – The upper bound of the filter when lo is not a string.

  • bkg_id (int or str, optional) – The filter will be applied to the associated background component of the data set if bkg_id is set. Only PHA data sets support this option; if not given, then the filter is applied to all background components as well as the source data.

See also

ignore

Exclude data from the fit.

sherpa.astro.ui.ignore2d

Exclude a spatial region from an image.

notice_id

Include data from the fit for a data set.

show_filter

Show any filters applied to a data set.

Notes

The order of ignore and notice calls is important.

The units used depend on the analysis setting of the data set, if appropriate.

To filter a 2D data set by a shape use ignore2d.

Examples

Ignore all data points with an X value (the independent axis) between 12 and 18 for data set 1:

>>> ignore_id(1, 12, 18)

Ignore the range up to 0.5 and 7 and above, for data sets 1, 2, and 3:

>>> ignore_id([1,2,3], None, 0.5)
>>> ignore_id([1,2,3], 7, None)

Apply the same filter as the previous example, but to data sets “core” and “jet”:

>>> ignore_id(["core","jet"], ":0.5,7:")
image_close()[source] [edit on github]

Close the image viewer.

Close the image viewer created by a previous call to one of the image_xxx functions.

See also

image_deleteframes

Delete all the frames open in the image viewer.

image_getregion

Return the region defined in the image viewer.

image_open

Start the image viewer.

image_setregion

Set the region to display in the image viewer.

image_xpaget

Return the result of an XPA call to the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Examples

>>> image_close()
image_data(id=None, newframe=False, tile=False)[source] [edit on github]

Display a data set in the image viewer.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_data_image

Return the data used by image_data.

image_close

Close the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_open

Open the image viewer.

image_source

Display the model for a data set in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the data in default data set.

>>> image_data()

Display data set 2 in a new frame so that the data in the current frame is not destroyed. The new data will be displayed in a single frame (i.e. the only data shown by the viewer).

>>> image_data(2, newframe=True)

Display data sets ‘i1’ and ‘i2’ side by side:

>>> image_data('i1')
>>> image_data('i2', newframe=True, tile=True)
image_deleteframes()[source] [edit on github]

Delete all the frames open in the image viewer.

Delete all the frames - in other words, images - being displayed in the image viewer (e.g. as created by image_data or image_fit).

See also

image_close

Close the image viewer.

image_getregion

Return the region defined in the image viewer.

image_open

Create the image viewer.

image_setregion

Set the region to display in the image viewer.

image_xpaget

Return the result of an XPA call to the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Examples

>>> image_deleteframes()
image_fit(id=None, newframe=True, tile=True, deleteframes=True)[source] [edit on github]

Display the data, model, and residuals for a data set in the image viewer.

This function displays the data, model (including any instrument response), and the residuals (data - model), for a data set.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

  • deleteframes (bool, optional) – Should existing frames be deleted? The default is True.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

image_close

Close the image viewer.

image_data

Display a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_resid

Display the residuals (data - model) for a data set in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the fit results - that is, the data, model, and residuals - for the default data set.

>>> image_fit()

Do not tile the frames (the three frames are loaded, but only the last displayed, the residuals), and then change the frame being displayed to the second one (the model).

>>> image_fit('img', tile=False)
>>> image_xpaset('frame 2')
image_getregion(coord='')[source] [edit on github]

Return the region defined in the image viewer.

The regions defined in the current frame are returned.

Parameters

coord (str, optional) – The coordinate system to use.

Returns

region – The region, or regions, or the empty string.

Return type

str

Raises

sherpa.utils.err.DS9Err – Invalid coordinate system.

See also

image_setregion

Set the region to display in the image viewer.

image_xpaget

Return the result of an XPA call to the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Examples

>>> image_getregion()
'circle(123,128,12.377649);-box(130,121,14,14,329.93142);'
>>> image_getregion('physical')
'circle(3920.5,4080.5,396.08476);-rotbox(4144.5,3856.5,448,448,329.93142);'
image_kernel(id=None, newframe=False, tile=False)[source] [edit on github]

Display the 2D kernel for a data set in the image viewer.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_kernel_image

Return the data used by image_kernel.

image_close

Close the image viewer.

image_data

Display a data set in the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_source

Display the model for a data set in the image viewer.

plot_kernel

Plot the 1D kernel applied to a data set.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

>>> image_kernel()
>>> image_kernel(2)
image_model(id=None, newframe=False, tile=False)[source] [edit on github]

Display the model for a data set in the image viewer.

This function evaluates and displays the model expression for a data set, including any instrument response (e.g. PSF or ARF and RMF) whether created automatically or with set_full_model.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_model_image

Return the data used by image_model.

image_close

Close the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model_component

Display a component of the model in the image viewer.

image_open

Open the image viewer.

image_source

Display the model for a data set in the image viewer.

image_source_component

Display a component of the source expression in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the model for the default data set.

>>> image_model()

Display the model for data set 2 in a new frame so that the data in the current frame is not destroyed. The new data will be displayed in a single frame (i.e. the only data shown by the viewer).

>>> image_model(2, newframe=True)

Display the models for data sets ‘i1’ and ‘i2’ side by side:

>>> image_model('i1')
>>> image_model('i2', newframe=True, tile=True)
image_model_component(id, model=None, newframe=False, tile=False)[source] [edit on github]

Display a component of the model in the image viewer.

This function evaluates and displays a component of the model expression for a data set, including any instrument response. Use image_source_component to exclude the response.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to display (the name, if a string).

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_model_component_image

Return the data used by image_model_component.

image_close

Close the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_source

Display the source expression for a data set in the image viewer.

image_source_component

Display a component of the source expression in the image viewer.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the full source model and then just the ‘gsrc’ component for the default data set:

>>> image_model()
>>> image_model_component(gsrc)

Display the ‘clus’ component of the model for the ‘img’ data set side by side without the with any instrument response (such as convolution with a PSF model):

>>> image_source_component('img', 'clus')
>>> image_model_component('img', 'clus', newframe=True,
...                       tile=True)
image_open()[source] [edit on github]

Start the image viewer.

The image viewer will be started, if found. Calling this function when the viewer has already been started will not cause a second viewer to be started. The image viewer will be started automatically by any of the commands like image_data.

See also

image_close

Close the image viewer.

image_deleteframes

Delete all the frames open in the image viewer.

image_getregion

Return the region defined in the image viewer.

image_setregion

Set the region to display in the image viewer.

image_xpaget

Return the result of an XPA call to the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

>>> image_open()
image_psf(id=None, newframe=False, tile=False)[source] [edit on github]

Display the 2D PSF model for a data set in the image viewer.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist.

See also

get_psf_image

Return the data used by image_psf.

image_close

Close the image viewer.

image_data

Display a data set in the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_source

Display the model for a data set in the image viewer.

plot_psf

Plot the 1D PSF model applied to a data set.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

>>> image_psf()
>>> image_psf(2)
image_ratio(id=None, newframe=False, tile=False)[source] [edit on github]

Display the ratio (data/model) for a data set in the image viewer.

This function displays the ratio data/model for a data set.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_ratio_image

Return the data used by image_ratio.

image_close

Close the image viewer.

image_data

Display a data set in the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_resid

Display the residuals (data - model) for a data set in the image viewer.

image_source

Display the model for a data set in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the ratio (data/model) for the default data set.

>>> image_ratio()
image_resid(id=None, newframe=False, tile=False)[source] [edit on github]

Display the residuals (data - model) for a data set in the image viewer.

This function displays the residuals (data - model) for a data set.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_resid_image

Return the data used by image_resid.

image_close

Close the image viewer.

image_data

Display a data set in the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_open

Open the image viewer.

image_ratio

Display the ratio (data/model) for a data set in the image viewer.

image_source

Display the model for a data set in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the residuals for the default data set.

>>> image_resid()

Display the residuals for data set 2 in a new frame so that the data in the current frame is not destroyed. The new data will be displayed in a single frame (i.e. the only data shown by the viewer).

>>> image_resid(2, newframe=True)

Display the residuals for data sets ‘i1’ and ‘i2’ side by side:

>>> image_resid('i1')
>>> image_resid('i2', newframe=True, tile=True)
image_setregion(reg, coord='')[source] [edit on github]

Set the region to display in the image viewer.

Parameters
  • reg (str) – The region to display.

  • coord (str, optional) – The coordinate system to use.

Raises

sherpa.utils.err.DS9Err – Invalid coordinate system.

See also

image_getregion

Return the region defined in the image viewer.

image_xpaget

Return the result of an XPA call to the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Examples

Add a circle, in the physical coordinate system, to the data from the default data set:

>>> image_data()
>>> image_setregion('circle(4234.53,3245.29,46.74)', 'physical')

Copy the region from the current frame, create a new frame displaying the residuals from data set ‘img’, and then display the region on it:

>>> r = image_getregion()
>>> image_resid('img', newframe=True)
>>> image_setregion(r)
image_source(id=None, newframe=False, tile=False)[source] [edit on github]

Display the source expression for a data set in the image viewer.

This function evaluates and displays the model expression for a data set, without any instrument response.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_source_image

Return the data used by image_source.

image_close

Close the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_model_component

Display a component of the model in the image viewer.

image_open

Open the image viewer.

image_source_component

Display a component of the source expression in the image viewer.

Notes

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the source model for the default data set.

>>> image_source()

Display the source model for data set 2 in a new frame so that the data in the current frame is not destroyed. The new data will be displayed in a single frame (i.e. the only data shown by the viewer).

>>> image_source(2, newframe=True)

Display the source models for data sets ‘i1’ and ‘i2’ side by side:

>>> image_source('i1')
>>> image_source('i2', newframe=True, tile=True)
image_source_component(id, model=None, newframe=False, tile=False)[source] [edit on github]

Display a component of the source expression in the image viewer.

This function evaluates and displays a component of the model expression for a data set, without any instrument response. Use image_model_component to include any response.

The image viewer is automatically started if it is not already open.

Parameters
  • id (int or str, optional) – The data set. If not given then the default identifier is used, as returned by get_default_id.

  • model (str or sherpa.models.model.Model instance) – The component to display (the name, if a string).

  • newframe (bool, optional) – Create a new frame for the data? If False, the default, then the data will be displayed in the current frame.

  • tile (bool, optional) – Should the frames be tiles? If False, the default, then only a single frame is displayed.

Raises

sherpa.utils.err.IdentifierErr – If the data set does not exist or a source expression has not been set.

See also

get_source_component_image

Return the data used by image_source_component.

image_close

Close the image viewer.

image_fit

Display the data, model, and residuals for a data set in the image viewer.

image_model

Display the model for a data set in the image viewer.

image_model_component

Display a component of the model in the image viewer.

image_open

Open the image viewer.

image_source

Display the source expression for a data set in the image viewer.

Notes

The function does not follow the normal Python standards for parameter use, since it is designed for easy interactive use. When called with a single un-named argument, it is taken to be the model parameter. If given two un-named arguments, then they are interpreted as the id and model parameters, respectively.

Image visualization is optional, and provided by the DS9 application [1]_.

References

1

http://ds9.si.edu/site/Home.html

Examples

Display the full source model and then just the ‘gsrc’ component for the default data set:

>>> image_source()
>>> image_source_component(gsrc)

Display the ‘clus’ and ‘bgnd’ components of the model for the ‘img’ data set side by side:

>>> image_source_component('img', 'clus')
>>> image_source_component('img', 'bgnd', newframe=True,
...                        tile=True)
image_xpaget(arg)[source] [edit on github]

Return the result of an XPA call to the image viewer.

Send a query to the image viewer.

Parameters

arg (str) – A command to send to the image viewer via XPA.

Returns

returnval

Return type

str

Raises
  • sherpa.utils.err.DS9Err – The image viewer is not running.

  • sherpa.utils.err.RuntimeErr – If the command is not recognized.

See also

image_close

Close the image viewer.

image_getregion

Return the region defined in the image viewer.

image_open

Create the image viewer.

image_setregion

Set the region to display in the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Notes

The XPA access point [1]_ of the ds9 image viewer lets commands and queries to be sent to the viewer.

References

1

http://ds9.si.edu/ref/xpa.html

Examples

Return the current zoom setting of the active frame:

>>> image_xpaget('zoom')
'1\n'
image_xpaset(arg, data=None)[source] [edit on github]

Return the result of an XPA call to the image viewer.

Send a command to the image viewer.

Parameters
  • arg (str) – A command to send to the image viewer via XPA.

  • data (optional) – The data for the command.

Raises
  • sherpa.utils.err.DS9Err – The image viewer is not running.

  • sherpa.utils.err.RuntimeErr – If the command is not recognized or could not be completed.

See also

image_close

Close the image viewer.

image_getregion

Return the region defined in the image viewer.

image_open

Create the image viewer.

image_setregion

Set the region to display in the image viewer.

image_xpaset

Send an XPA command to the image viewer.

Notes

The XPA access point [1]_ of the ds9 image viewer lets commands and queries to be sent to the viewer.

References

1

http://ds9.si.edu/ref/xpa.html

Examples

Change the zoom setting of the active frame:

>>> image_xpaset('zoom 4')

Overlay the coordinate grid on the current frame:

>>> image_xpaset('grid yes')

Add the region file ‘src.reg’ to the display:

>>> image_xpaset('regions src.reg')

Create a png version of the image being displayed:

>>> image_xpaset('saveimage png /tmp/img.png')
int_proj(par, id=None, otherids=None, replot=False, fast=True, min=None, max=None, nloop=20, delv=None, fac=1, log=False, numcores=None, overplot=False)[source] [edit on github]

Calculate and plot the fit statistic versus fit parameter value.

Create a confidence plot of the fit statistic as a function of parameter value. Dashed lines are added to indicate the current statistic value and the parameter value at this point. The parameter value is varied over a grid of points and the free parameters re-fit. It is expected that this is run after a successful fit, so that the parameter values identify the best-fit location.

Parameters
  • par – The parameter to plot.

  • id (int or str, optional) – The data set that provides the data. If not given then all data sets with an associated model are used simultaneously.

  • otherids (sequence of int or str, optional) – Other data sets to use in the calculation.

  • replot (bool, optional) – Set to True to use the values calculated by the last call to int_proj. The default is False.

  • fast (bool, optional) – If True then the fit optimization used may be changed from the current setting (only for the error analysis) to use a faster optimization method. The default is False.

  • min (number, optional) – The minimum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • max (number, optional) – The maximum parameter value for the calculation. The default value of None means that the limit is calculated from the covariance, using the fac value.

  • nloop (int, optional) – The number of steps to use. This is used when delv is set to None.

  • delv (number, optional) – The step size for the parameter. Setting this over-rides the nloop parameter. The default is None.

  • fac (number, optional) – When min or max is not given, multiply the covariance of the parameter by this value to calculate the limit (which is then added or subtracted to the parameter value, as required).

  • log (bool, optional) – Should the step size be logarithmically spaced? The default (False) is to use a linear grid.

  • numcores (optional) – The number of CPU cores to use. The default is to use all the cores on the machine.

  • overplot (bool, optional) – If True then add the data to an existing plot, otherwise create a new plot. The default is False.

See also

conf

Estimate patameter confidence intervals using the confidence method.

covar

Estimate the confidence intervals using the covariance method.

get_int_proj

Return the interval-projection object.

int_unc

Calculate and plot the fit statistic versus fit parameter value.

reg_proj

Plot the statistic value as two parameters are varied.

Notes

The difference to int_unc is that at each step, a fit is made to the remaining thawed parameters in the source model. This makes the result a more-accurate rendering of the projected shape of the hypersurface formed by the statistic, but the run-time is longer than, the results of int_unc, which does not vary any other parameter. If there are no free parameters in the source expression, other than the parameter being plotted, then the results will be the same.

Examples

Vary the gamma parameter of the p1 model component for all data sets with a source expression.

>>> int_proj(p1.gamma)

Use only the data in data set 1:

>>> int_proj(p1.gamma, id=1)

Use two data sets (‘obs1’ and ‘obs2’):

>>> int_proj(clus.kt, id='obs1', otherids=[