After the patch went up for r243, I noticed a few issues with the dark and
flat generation in FakeCamera:
The dark generation was giving values that were too high because I forgot
to divide by e-/ADU.
You cannot place a standard random function into numpy.fromfunction()
because when using lambda, Python optimizes and you get a single random
value applied everywhere rather than different random values applied at
various locations.
Log messages are now more descriptive and errors are logged in a more
discrete manner specific to the failure that causes them.
There was an issue with the telescope/dome alignment code for AZ near 0/360.
Simulated flats weren't being produced if the dome or telescope weren't
available; instead, you just got a 100x100 set of zeros. Now, flats are
attempted if the shutter is open and either the telescope or the dome
aren't available.
Otherwise, everything looks good. Good catch on removing the for loops.
Please see the attached patch for fixes.
Original issue reported on code.google.com by laughing...@gmail.com on 7 Jun 2008 at 3:03
Original issue reported on code.google.com by
laughing...@gmail.com
on 7 Jun 2008 at 3:03Attachments: