new version for the 20131202 release?

[scottransom]

Hi @nirinA,

Thanks a ton for these wrappers. They will likely be super useful for some pulsar timing code that I'm helping to develop.

Do you intend to keep updating the code? The new SOFA release from 20131202 contains several new astrometry routines which would be useful to have...

If you are not intending to continue updating, would you be willing to release your scripts that auto-generated the wrapper code? I'm assuming that that is at least partially what you have?

Thanks again. Really useful stuff.

Cheers,

Scott

PS: BTW, in order to get the current module to compile, all of the constants defined with PyFloat_FromDouble(DS2R) and similar need to be changed to PyFloat_FromDouble(ERFA_DS2R)...

[nirinA]

Le Sun, 22 Dec 2013 09:24:17 +0300, Scott Ransom
notifications@github.com a écrit:

Hi @nirinA,

hi Scott,

Thanks a ton for these wrappers. They will likely be super useful for
some pulsar timing code that I'm helping to develop.

i'm glad to hear the wrappers may be useful, and i'll be really
happy if i can add my little help to develop those pulsar timing
code.

Do you intend to keep updating the code? The new SOFA release from
20131202 contains several new astrometry routines which would be useful
to have...

yes, i've started to add the new routines from the latest SOFA release.
i hope i'll have some free time soon to complete the package.
i'll try to commit the first routines i wrote this afternoon.

If you are not intending to continue updating, would you be willing to
release your scripts that auto-generated the wrapper code? I'm assuming
that that is at least partially what you have?

yes, the extension module is partially auto-generated. the scripts
intent to write a correct Python extension module. you still have
to write by hand the body of the C functions.
i'll also put them on github soon, after a bit polishing. they are
rather ugly for now.

Thanks again. Really useful stuff.

Cheers,

Scott

PS: BTW, in order to get the current module to compile, all of the
constants defined with PyFloat_FromDouble(DS2R) and similar need to be
changed to PyFloat_FromDouble(ERFA_DS2R)...

yes, the new wrapper should take care of that.
thanks you very much,

best regards,

nirinA

---

Reply to this email directly or view it on GitHub:
#2

[nirinA]

i added 9 new routines from the astrometry tools. starting with the aper routine, i think the PyStructSequence is not
the right way to handle ASTROM and LBODY struct and one needs to build a new type object, which is a bit more complex.

[scottransom]

Fantastic! Thanks for that. I'm still getting errors with the constants during compile, though....

And yes, the ASTROM and LBODY structs seem a bit more complex than what was in there before.

The timing code will use parts of your wrappers almost certainly. We are in the process of testing some things with the "standard" pulsar timing software TEMPO2 now. The code is very early, but if you want to poke at it, it is here:
https://github.com/scottransom/PINT

Thanks again! And Happy Holidays!

[nirinA]

Le Sun, 22 Dec 2013 20:04:40 +0300, Scott Ransom
notifications@github.com a écrit:

Fantastic! Thanks for that.

you are welcome!

I'm still getting errors with the constants during compile, though....

that's weird, i compiled the package with gcc-4.8.2, Python-3.4.0b1
and it succeeded.

And yes, the ASTROM and LBODY structs seem a bit more complex than what
was in there before.

The timing code will use parts of your wrappers almost certainly. We
are in the process of testing some things with the "standard" pulsar
timing software TEMPO2 now. The code is very early, but if you want
to poke at it, it is here:
https://github.com/scottransom/PINT

i just create a fork into my repo and in the process
to download TEMPO2.
i'm interested to help with unittest, but need to learn
more about what is needed to be done.

Thanks again! And Happy Holidays!

with best regards,

nirinA

[nirinA]

i push a new version of the package. the extension module was reworked. finally, i keep using structseq. just create a pure Python module for routines using LDBODY.
all routines from the latest SOFA release are added. they need some more check and polishing, though.

[nirinA]

hi @scottransom
i close this issue and let you know that a new wrapper using numpy is available at

https://github.com/nirinA/erfa_numpy

some functions related to vector/matrix manipulation are still to be added.