search

esheldon / psfsim_config

Configuration files for the psf simulations
0 stars 0 forks source link

Choose number of images for a first big run #8

Open esheldon opened 8 years ago

esheldon commented 8 years ago

@rmjarvis what would be a good number of images to test rho statistics?

rmjarvis commented 8 years ago

Something equivalent to the test bed would probably be good for now. Maybe of order 100 exposures?

esheldon commented 8 years ago

I'm running one with 20,000 now, equivalent to ~339 exposures of 59 ccds

I considered that small; I thought you would need a much larger run for rho stats.

rmjarvis commented 8 years ago

We'll see, but I suspect that will be good to look at for now. If we need to, we can go larger of course, but I think this is plenty large enough to learn something from.

esheldon commented 8 years ago

OK, they are appearing here at BNL: 

/gpfs/mnt/gpfs01/astro/workarea/esheldon/lensing/des-lensing/psfsim/v005/output
esheldon commented 8 years ago

v005 finished. Maybe worth taking a look despite the bugs.

You can get file names using the psfsim package.

import psfsim
run='v005'
index=3421
imfile=psfsim.files.get_stars_file(run, index)
truth_file=psfsim.files.get_stars_truth_file(run, index)
esheldon commented 8 years ago

Need to set this env. variable

PSFSIM_DIR=/gpfs/mnt/gpfs01/astro/workarea/esheldon/lensing/des-lensing/psfsim
esheldon commented 8 years ago

That bug in gsparams meant that folding_threshold: 1.e-3 did not get set. The code now runs 3 times slower, but from the comment it seems this is absolutely needed

esheldon commented 8 years ago

A new version v006 has run, with bug fixes in place.