Last modified: 27 April 2021

How does Sherpa calculate errors on PHA data values given as a rate rather than counts?

When a PHA file with a RATE column is read into Sherpa, Sherpa automatically multiplies the count rate in that column by the exposure time read from the EXPOSURE keyword in that file (be it a source or background spectrum), converting the rate to counts. This is because PHA data are fit in channel space in Sherpa, i.e., counts as a function of channels, in order to use Poisson statistics. Therefore, the rate is converted to counts in this instance, and the statistical errors are computed on the counts values using the chosen fit statistic.