QUBIC

Minutes:

• FMM comments (Steve and others):
  - get_dict is in five different places, but it should be initiated before creating the objects
  - use dictionnary everywhere and it should be always the first argument, and some parts of the dictionnary can be modified in the init functions
  - hierarchy in the qubic_dict doesn't seem necessary, but maybe we keep it for backward compatibility
  - we should use only one type of file to create the dictionnaries
  - maybe have two dictionnaries? One for the intrument and one for the analysis
  - seperate simulation from reconstruction, but need to avoid writing TODs (for now) while creating a way to save/read data (fits files?)
  - pbs are in
  qubic/lib/Instrument/Qnoise.py (it's in init)
  qubic/lib/MapMaking/ComponentMapMaking/preset/preset_qubic.py
  qubic/lib/MapMaking/FrequencyMapMaking/Qfmm.py
  qubic/lib/MapMaking/Qatmosphere_2d.py
  qubic/lib/Qspectra.py
  but also in Qfmm (careful get_dict is used here), ...
  - add the parameters of the intrument used in the simulation in the .pkl file saved + date and other stuff (?) because currently only the one used in the mapmaking is saved (?)
  - what is the "tod" key in Steve's maps? When was it added? (ask Tom)

• Moon (Alexandre):
  - showed the stacked map of the Moon
  --> see these presentations: http://qubic.in2p3.fr/wiki/pmwiki.php/SimulationsWorkingGroup/Mapmaking#Moon_Maps
  --> maybe harder selection of TES
  --> got to check in details how Giuseppe got the spectrum
  --> check if the subpeaks are at the same place in "Moon space", or if we should change space (maybe use SimulateMoonTOD-RealData.ipynb)
  --> add an unknown intercalibration between TES during the stacking (use Intercal Moon Maps.ipynb)

• 1D QUBIC (JC):
  - effect of realism (pixelisation of H) and GPA (gaussian approximation) might not be negligible