Recreating some IRAF functionality?

Hi there,

I have some code that relies on IRAF (e.g. median, imcombine, blkrep, ...). I'm planning on rewriting it to be pure python. I looked through astropy and the affiliated packages to see if python versions of the IRAF commands already existed and didn't find anything. Since this package seemed to be the place where IRAF functions would most likely be recreated I thought I'd double check here.

If indeed these functions don't exist, would it be something to include in the ccdproc package? For example, I was thinking of writing a wrapper for the scipy.signal.medfilt to have it work more like the IRAF task median (I did see that there is a cosmicray_median function in this package, maybe some functionality could be reused?).

I'm not very familiar with IRAF (which is why I want to change the code to use python instead) so I'm not sure if the reason these functions don't exist in astropy/the affiliated packages is because people don't like them for some reason. Please let me know if that's the case.

Thanks,
Alexa

It would be great for you to contribute whatever you have and would be wonderful to include it. The reason some of the functionality may not exist is typically because we have not had time to implement it and so if there is stuff you would like to see in ccdproc, please let us know. So it's not at all that we do not like them, but either we haven't needed them or just have not implemented them yet. So having your help and contributes would be a big help.

My suggestion for how best to proceed is to open an issue for each feature that you would like to see but can't find in ccdproc. Somethings do exists in other forms and just are currently not wrapped in ccdproc and so those might be quite easy to include (like the median example you give--it would be great to include that as a function in ccdproc). We also might be able to suggest some other places that things do show up or where there is overlap or the best way to proceed (which in some cases might be implementing the code and in others it might be better or more visible documentation).

Nonetheless, it would be great to have your contributions and thank you for your interest!

@justincely

For imcombine, STScI has some kind of Python implementation already. See https://github.com/spacetelescope/stsci.image/blob/master/stsci/image/numcombine.py . We do plan to consolidate such implementations currently scattered across the board somewhere easier to find. c/c @saraogaz, @justincely

I was pretty sure we also re-implemented blkrep in Python before but I can't find it now... @larrybradley , was it absorbed into stats or photutils?

EDIT: Okay, we didn't re-implement blkrep but rather blkavg. Not sure if that one made it into official STScI distribution, but I have a copy at https://github.com/pllim/playpen/blob/master/blkavg_rewrite.py (still need some work to optimize it but I think it works)

@pllim

@pllim blkrep could be astropy.nddata.utils.block_replicate and astropy.nddata.utils.block_reduce or ccdproc.rebin (this might be subject to change because it doesn't do what people expect #388).

@pllim

@pllim Yes, I added block_replicate and block_reduce to astropy. block_reduce is flexible in that it can apply any user-input function to the data. The default is func=np.sum, which is equivalent to ~~blkrep~~ block summing. The equivalent of blkavg would be to input func=np.mean (or e.g. np.ma.mean).

Sorry, blk_rep is block_replicate not block_reduce.....

Also imcombine is probably already implemented in ccdproc: low-level interface Combiner and high-level combine

@AlexaVillaume

Hi @AlexaVillaume,
This is super related to a project we're working on right now at STScI to replace needed IRAF tasks and functionality. Can I ask what part of the IRAF functionality you want that's not part of scipy.signal.medfilt ? This could easily be something that goes on our to-do list...

@pllim

Hi @SaraOgaz,

My thought was to make a wrapper to the scipy functionality so that one could just pass a name of a fits file, like in IRAF. Then also to work properly over MEFs, which I don't think the IRAF version does. Is this the same project that @pllim brought up?

@AlexaVillaume

@AlexaVillaume , yes, the very same project.

What might also be of help is the following project:
https://stak-notebooks.readthedocs.io/en/latest/

We are going to close this issue for right now, but please open individual issues if there is a specific functionality to implement.

Navigation Menu

Repository navigation

Recreating some IRAF functionality? #370

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Recreating some IRAF functionality? #370

Description

Activity

crawfordsm commented on Jul 7, 2016

pllim commented on Sep 8, 2016

pllim commented on Sep 8, 2016

MSeifert04 commented on Sep 8, 2016

larrybradley commented on Sep 8, 2016

larrybradley commented on Sep 8, 2016

MSeifert04 commented on Sep 8, 2016

SaOgaz commented on Oct 19, 2016

AlexaVillaume commented on Oct 20, 2016

pllim commented on Oct 20, 2016

crawfordsm commented on Dec 7, 2018

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions