Tasks for Rutgers hackathon

LSSTDESC / DEHSC_LSS

Scripts and workflow relevant to HSC data

1 stars 3 forks source link

Tasks for Rutgers hackathon #21

Closed damonge closed 6 years ago

damonge commented 6 years ago

[x] Wiki with cuts/specs/plots (#19)
[x] Implement standard pdf stacking in cat_sampler.py (#17)
[x] Understand PIT and test it to obtain calibrated N(z)s (#17)
[x] Decide tomography scheme (#16)
[x] Start implementing the 2-pt correlation function estimation (#22)
[x] Make full pipeline run and obtain power spectra
[x] If we get here we're superheroes, in which case we get started with likelihood (#20)
[ ] N(z) estimates from COSMOS 30-band photometry (#18)

damonge commented 6 years ago

And @humnaawan ticks off the first task of the hackathon! (#16)

adam-broussard commented 6 years ago

I was able to implement PDF stacking for the N(z)'s in cat_sampler.py. These plots show the similar N(z) distributions after calculating them using point redshifts and the PDFs.

pdfstack z_mc

adam-broussard commented 6 years ago

We ran the pipeline to completion and generated variance-weighted angular power spectrum plots. Note that these are weighted by survey area rather than the actual variance, so the normalization is likely bonkers.

powerspectrum

damonge commented 6 years ago

7/8 tasks done (including "being superheroes"). Not too bad guys!!!