Deciding on storage format for shear and convergence maps

LSSTDESC / WLMassMap

DESC utility for weak lensing mass-mapping

BSD 3-Clause "New" or "Revised" License

2 stars 0 forks source link

Deciding on storage format for shear and convergence maps #26

Open EiffL opened 4 years ago

EiffL commented 4 years ago

@maho3 realized that we had some inconsistencies in the way we are storing the shear and mass maps compared to TXPipe and we should convergence on the same choices. Here are a few options;

[ ] Storing the maps as astropy table in our own FITS format (what we were doing)
[ ] Storing the maps as HDF5 astropy table (what we are doing now)
[ ] Check what format is used by NaMaster or TXpipe (what we should be doing)

fjaviersanchez commented 4 years ago

@EiffL, NaMaster uses HEALPix maps and flat-sky maps more information about how to create these here. @erykoff, @esheldon and I were also working on a sparse version of healpix, healsparse that you may want to consider (shameless plug 😛 )

joezuntz commented 4 years ago

I'd like to move the main TXPipe over to Healsparse, so if you get a chance to use it in the mass mapping work then that would be a v useful reference.

EiffL commented 4 years ago

Healparse it is !

patricialarsen commented 4 years ago

@EiffL Can you please confirm that you mean the shear and convergence maps specifically in the Mass Mapping work, and not the inputs? (these are currently in Healpix fits file format)