For a typical XMM observation, there is data from three instruments. We will assume that the spectra have, as per ยง5.14, been grouped, and the names of the redistribution file (.rmf), the ancillary response file (.arf), and background file (-bkg.pi) have already been added to the header. Then we can import the data:
data 1:1 mos1S001-fovt-grp.pi
ig 1:0.0-0.3,11.0-**
data 2:2 mos2S002-fovt-grp.pi
ig 2:0.0-0.3,11.0-**
data 3:3 pnS003_0-fovt-grp.pi
ig 3:0.0-0.3,2.0-**
data 4:4 pnS003_4-fovt-grp.pi
ig 4:0.0-1.0,11.0-**
ig 4:7.2-9.2
data 5:5 rosat.pi
ig bad
Both of the MOS spectra are used for energies in E. Note that we are importing two different versions of the pn data, one filtered to contain only the PATTERN==0 events, and the other filtered to include PATTERN4. The former is used only for E while the latter is used for E. For the pn data we exclude 7.2 to 9.2 keV in order to avoid fitting the several instrumental lines in that range. We have also included a spectrum from ROSAT in order to better constrain the lower end of the fit.