These are chat archives for astropy/astropy

3rd
May 2017
Gregory Green
@gregreen
May 03 2017 02:40
@eteq I tested the SkyCoord reshaping timings again, using the latest dev version of astropy, and it looks like the overhead has gone down to 2.5 ms (from 30 ms previously).
However, I have another question about speed in astropy.
Gregory Green
@gregreen
May 03 2017 02:50
If I have a large number of equatorial coordinates that I'd like to convert to alt/az at a range of times, it can take a large amount of time (in my case, on the order of 0.5 ms per coordinate or time, which can add up). Is it safe to assume that the bulk of the time is consumed by doing the lookups/transformations relating to Earth's instantaneous position and the instantaneous position of the location on Earth?
Gregory Green
@gregreen
May 03 2017 03:02
Doing the (RA,Dec) -> (Alt,Az) transformation for a bunch of coordinates at the same time doesn't seem to speed up the transformation, though.
Gregory Green
@gregreen
May 03 2017 10:04
The reason why I'm raising this is because I'm trying out astroplan. Computing the constraints for 50 fixed targets (which involves checking if the Sun is down, if the Moon is within 30 degrees, and computing airmasses) takes 16 seconds on my computer. This basically makes observing strategy optimization infeasible. Any ideas about where the time is being spent?