group, bkg_id=None)

Turn on the grouping for a PHA data set.

A PHA data set can be grouped either because it contains grouping information 1, which is automatically applied when the data is read in with load_pha or load_data, or because the group set of routines has been used to dynamically re-group the data. The ungroup function removes this grouping (however it was created). The group function re-applies this grouping. The grouping scheme can be changed dynamically, using the group_xxx series of routines.

  • 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.

  • bkg_id (int or str, optional) – Set to group the background associated with the data set.


sherpa.utils.err.ArgumentErr – If the data set does not contain a PHA data set.

See also


Fit one or more data sets.


Adaptively group to a minimum number of counts.


Adaptively group to a minimum signal-to-noise ratio.


Group into a fixed number of bins.


Group into a minimum number of counts per bin.


Group into a minimum signal-to-noise ratio.


Group into a fixed bin width.


Apply a set of grouping flags to a PHA data set.


Apply a set of quality flags to a PHA data set.


Turn off the grouping for a PHA data set.


PHA data is often grouped to improve the signal to noise of the data, by decreasing the number of bins, so that a chi-square statistic can be used when fitting the data. After calling group, anything that uses the data set - such as a plot, fit, or error analysis - will use the grouped data values. Models should be re-fit if group is called; the increase in the signal of the bins may mean that a chi-square statistic can now be used.

The grouping is implemented by separate arrays to the main data - the information is stored in the grouping and quality arrays of the PHA data set - so that a data set can be grouped and ungrouped many times, without losing information. The group command does not create this information; this is either created by modifying the PHA file before it is read in, or by using the group_xxx routines once the data has been loaded.

The grouped field of a PHA data set is set to True when the data is grouped.



Arnaud., K. & George, I., “The OGIP Spectral File Format”,


Group the data in the default data set:

>>> group()
>>> get_data().grouped

Group the first background component of the ‘core’ data set:

>>> group('core', bkg_id=1)
>>> get_bkg('core', bkg_id=1).grouped

The data is fit using the ungrouped data, and then plots of the data and best-fit, and the residuals, are created. The first plot uses the ungrouped data, and the second plot uses the grouped data.

>>> ungroup()
>>> fit()
>>> plot_fit_resid()
>>> group()
>>> plot_fit_resid()