AlistairSymonds

icon indicating a website link https://www.linkedin.com/in/alistair-symonds/ icon indicating an email [email protected]

icon location Sydney icon location GPU ASIC design by day, all sorts of other tinkering at night. All this is personal and I don't represent any company in any way here.

Results 6 comments of AlistairSymonds

Q2d_nm_c_to_a_b needs a non-zero positive and negative m to work

> Since I am very bad at math and the creator never wrote down how to do do it, there is not a Q2D_der function. Right there with you with...

Q2d_nm_c_to_a_b needs a non-zero positive and negative m to work

> Realistically, the Q2d_nm_c_to_a_b is probably a mistake, since for example n=4,m=0 is spherical in Zernikes, but it's n=0,m=0 that is most similar in Qs. And for "primary astigmatism" with...

Make sure reprojection routines work with large images

Just came across this issue when trying to put together a [silly large widefield mosaic](https://github.com/AlistairSymonds/astro-project-patchwork) which even memory mapped arrays failed at halfway through the pix2pix roundtrip. I basically just...

Make sure reprojection routines work with large images

Definitely going to give it a shot, I'm just trying to think about how to handle memory mapped arrays since this way memory mapped and multiprocessing support could be added...

Make sure reprojection routines work with large images

I've done a version for the non-healpix functions in the utils package, first off I'm not sure if this the correct place it makes sense to me https://github.com/AlistairSymonds/reproject/blob/dev_reproject_block/reproject/utils.py It takes...

Make sure reprojection routines work with large images

@astrofrog I'll give you another ping because this is pretty much done functionality wise, I didn't run out of memory when dealing with 100k * 200k output projections and the...