======================================================================
ERROR: test_main_defaults (desisim.test.test_pixsim.TestPixsim)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/global/common/software/desi/cori/desiconda/20200801-1.4.0-spec/code/desisim/master/py/desisim/test/test_pixsim.py", line 199, in test_main_defaults
desisim.scripts.pixsim.main(pixsimargs)
File "/global/common/software/desi/cori/desiconda/20200801-1.4.0-spec/code/desisim/master/py/desisim/scripts/pixsim.py", line 124, in main
simulate_exposure(args.simspec, args.rawfile, cameras=args.cameras,
File "/global/common/software/desi/cori/desiconda/20200801-1.4.0-spec/code/desisim/master/py/desisim/pixsim.py", line 158, in simulate_exposure
image, rawpix, truepix = simulate(camera, simspec, psf, comm=comm_node, preproc=False, cosmics=cosmics, **kwargs)
File "/global/common/software/desi/cori/desiconda/20200801-1.4.0-spec/code/desisim/master/py/desisim/pixsim.py", line 331, in simulate
photpix2raw(pix[0:ny, 0:nx], gain, rdnoiseA,
File "/global/common/software/desi/cori/desiconda/20200801-1.4.0-spec/code/desisim/master/py/desisim/pixsim.py", line 516, in photpix2raw
img = np.random.poisson(img) + noise
File "mtrand.pyx", line 3569, in numpy.random.mtrand.RandomState.poisson
File "_common.pyx", line 824, in numpy.random._common.disc
File "_common.pyx", line 621, in numpy.random._common.discrete_broadcast_d
File "_common.pyx", line 357, in numpy.random._common.check_array_constraint
ValueError: lam < 0 or lam contains NaNs
I'm pretty sure this was caused by specter commit #6a0d107 which now allows PSF projection of negative flux, which was needed for proper normalization when projecting ringing deconvolved real data spectra for studying the 2D model. i.e. now it is the responsibility of the input spectrum to not include negatives if they aren't supposed to be there. I haven't tracked why desisim has negative flux resulting in negative pixels resulting in failing poisson (or if that is actually what is happening).
Test failure:
I'm pretty sure this was caused by specter commit #6a0d107 which now allows PSF projection of negative flux, which was needed for proper normalization when projecting ringing deconvolved real data spectra for studying the 2D model. i.e. now it is the responsibility of the input spectrum to not include negatives if they aren't supposed to be there. I haven't tracked why desisim has negative flux resulting in negative pixels resulting in failing poisson (or if that is actually what is happening).