Update existing modules to astropy - #8
Conversation
|
I have started to go through this (but not pushed to github yet). My goal for now is to do the "minimum possible replacement", i.e. replace I have one problem and I would like to hear your opinion before I spend too much time implementing changes. The stuff that deals purely with tables is already in @astrofrog 's chapter So, I suggest to remove http://python4astronomers.github.io/files/atpy.html . What do you think? |
|
I agree we should drop ATpy as the primary package being used to read files. But it is certainly worth keeping a paragraph that describes ATpy, gives a link, and talks about additional functionality like the SQL interface. Will a We should also think about whether it makes sense to migrate some of the ATpy SQL interface to astropy. |
|
Maybe we go the other way, keep the atpy chapter for now, but add a big note on the top, noting that astronomy tables are probably better done in astropy.Table for now, but some functionality has not been migrated yet? @astrofrog What do you think? |
|
@taldcroft - I think Regarding the SQL stuff, I've been thinking that whatever is in ATpy that we decide not to migrate to Astropy core could end up in an affiliated package, so that there is no duplication, then ATpy itself can be deprecated. |
|
@hamogu - well, it's true that the lack of support for FITS tables with Table.read in Astropy 0.2 is kind of a pain... Maybe we can wait until 0.3 to update this? |
|
A rendered version of the suggested changes can be seen at |
There was a problem hiding this comment.
pip install --upgrade ...
|
On a related note you should search / replace |
There was a problem hiding this comment.
Best to use the "official" package version of all the readers, i.e. astropy.io.ascii.AASTex instead of astropy.io.ascii.latex.AASTex. Users aren't supposed to know where the AASTex class really comes from.
There was a problem hiding this comment.
I have changed that in python4astronomers, but since we link to the auto-doc generated documentation, you will see the place where this object really comes from and not the "lazy user way" to import it., once you click the link, e.g. in http://docs.astropy.org/en/v0.2.1/_generated/astropy.io.ascii.cds.Cds.html#astropy.io.ascii.cds.Cds
the docs say
class astropy.io.ascii.cds.Cds(readme=None)
Bases: astropy.io.ascii.core.BaseReader
(In fact, that's where I copied the links).
|
BTW, in answer to previous discussion, I would be in favor of gradually migrating python4astronomers to astropy based on the currently released functionality. So these changes are good, pending the minor comments I had. |
|
If there are no more comments, then I will merge this mid-week and also push to the live website. |
3 participants
While some of the chapters are timeless and still valid (python, numpy, scipy, matplotlib), some should be rewritten for use with astropy instead of deprecated modules that are now superceeded by
astropy(asciitable,pyfits, ...)