Add obj.makePhot method #1078
Add obj.makePhot method #1078
Conversation
That was originally my plan, but with all the maxN and resume stuff, if seemed a little too much to untangle cleanly. This seemed simpler, and the unit tests make sure they are equivalent for their overlapping use cases. |
galsim/gsobject.py
Outdated
| @@ -2067,6 +2067,101 @@ def _calculate_nphotons(self, n_photons, poisson_flux, max_extra_noise, rng): | |||
| return iN, g | |||
|
|
|||
|
|
|||
| def makePhot(self, n_photons=0, rng=None, max_extra_noise=0., poisson_flux=None, | |||
| sensor=None, surface_ops=(), maxN=None, orig_center=PositionI(0,0), | |||
There was a problem hiding this comment.
Sensor isn't in the parameter list.
There was a problem hiding this comment.
Thanks. Also maxN. Fixed.
galsim/gsobject.py
Outdated
| Returns: | ||
| - a `PhotonArray` with the data about the photons. | ||
| """ | ||
| from .sensor import Sensor |
There was a problem hiding this comment.
Is this necessary in this function?
There was a problem hiding this comment.
Thanks. No not needed here.
These tests check that drawPhot and makePhot produce the same exact photons. So the first time, for drawPhot, I call rng.duplicate() to fork the RNG sequence. Then when I do it again with makePhot, the rng runs through the exact same sequence again, so should produce the same photons. |
rmjarvis
added
base classes
This PR adds a way to draw photons without actually accumulating them onto an image.
It is functionally equivalent to
_, photons = obj.drawPhot(im, ...)but without requiring the image.@cwwalter