Tests of the new aeff phi dependence

Phi dependence is completely coded. Changes were required in the packages irfs/handoff_response, Likelihood, map_tools. All parts were separately tested, here we perform an overall check before the tools are considered ready for science analysis.

A new "P6_V5_DIFFUSE" irf was generated, with phi dependence in the aeff (see the attachements for a tarball). The pruned goodEvents dataset used was the same used to generate "P6_V3_DIFFUSE", so the 2 irf sets should be directly comparable.

Aeff comparison

First, pyIrfLoader was used to compare the olf aeff with the new, averaged over all phi. This should make the 2 results identical. Let's see:

Again, second plot is the aeff averaged over phi, the average is calculated by the script I used to make the plots. The P6V3 plot is the usual one.
Same for acceptance:

Obssim simulation

I simulated 2 datasets, one with P6v3 and the other one with P6v5. I'll analyze them separately with the same irf set used for the simulation. 1 week, point source at (RA,DEC)=(60,0) spectrum is a power law, index -2. All is ok here: see the PHI-THETA plots below.

Livetime

Livetime cubes now contain the usual entries in the first (40) bins for eack healpix pixel. After that the same cumulation is repeated in several phi bins, the number of phi bins can be specified with the paramater phibins. Let's see an example with 10 phi bin, I look at the first pixel in the cube:

Looks plausible. Zooming in to select only the total accumulation in the P6V5 plot on the right, this is found to be identical to the plot of P6V3:

Exposure map

Exposure map, 30 degrees in radius, 60x60 bins, 35 energy bins from 100 MeV to 200 GeV.
The plots look ok, I write down some numbers for the exposure in the source position:

The new exposure seems to be too high by about a 10%.

Likelihood fit

Fit results:

Not surprisingly the fitted spectrum for P6_v5 is too low by about 10%. Spectral index is correct though.