search

astropy / coordinates-benchmark

Accuracy Benchmarks of different coordinate conversion codes. Maintainers: @astrojuanlu, @astrofrog, and @cdeil
https://www.astropy.org/coordinates-benchmark/summary.html
11 stars 12 forks source link

SLALIB results for FK5 -> FK4 are not consistent with PyAST #21

Closed astrofrog closed 11 years ago

astrofrog commented 11 years ago

The SLALIB results for FK5 -> FK4 are off by 0.2" arcseconds (but the FK4 -> FK5 results agree). I wanted to try and debug this, but I can't get pyslalib to build correctly on my machine (something to do with the default architecture of the fortran compiler). One possible bug I can see is that the transformations should be at epoch J2000, but the following is B2000:

        # FK5 -> FK4 at BEPOCH 2000 assuming no proper motion or parallax
        lon, lat = slalib.sla_fk54z (lon, lat, 2000)[:2]

Maybe that line should be:

        lon, lat = slalib.sla_fk54z (lon, lat, 2000.0012775136652)[:2]

I can't test to see if it improves things though. @scottransom and @cdeil - any ideas? Does the above fix things?

scottransom commented 11 years ago

This isn't easy at all for me to test for the next couple days. Maybe Christoph can take a quick look?

astrofrog commented 11 years ago

I got mixed up, it's the FK4 -> FK5 results that are offset:

pyast      pyslalib   fk4        fk5            0.217934     0.196144     0.249062     0.055751
astrofrog commented 11 years ago

It might be something to do with the fact the observation epoch should be J2000, but I'm not sure how that should be specified.

astrofrog commented 11 years ago

I figured it out - the epoch in SLA_FK45Z should be the observation epoch, not the equinox. I'll fix directly in master.

astrofrog commented 11 years ago

Fixed in 1633675dfbb19d73264d8fdea99215b33d7e1002