[SciPy-Dev] IDL .sav file reader for scipy.io

Thomas Robitaille thomas.robitaille@gmail....
Sun Jul 25 11:16:53 CDT 2010


On Jul 25, 2010, at 11:58 AM, Pauli Virtanen wrote:

> Sun, 25 Jul 2010 09:10:36 -0400, Thomas Robitaille wrote:
>> A couple of weeks ago, I wrote a message to this list ('potential
>> scikits package') asking whether a package I have developed to read IDL
>> .sav files would be useful to have as a scikits package, but several
>> people suggested it should go into scipy.io directly.
> 
> +1
> 
> Looks good to me. Has tests and seems to work, so it should go in for 0.9 
> IMHO.

Thanks for looking over it!

> Some comments:
> 
> - verbose=False should be the default; a function in a library shouldn't
>  print anything unless explicitly asked to.

Ok - I will change that when including it in scipy.io (I decided to leave it as it is in IDLSave since that has been the default since the start).

> - What's the motivation for the `python_dict` option?

I added that in case someone wanted a proper dictionary rather than the overloaded dictionary type with attribute access. It probably won't get used much, but I just put it there in case. Do you think it should be removed?

> - What's the motivation for the `uncompressed_file_name` option?
>  For working with very large files?

Yes - in some cases it might be desirable to not have to read from the compressed file each time, and this can also be used if the uncompressed file should be expanded to a scratch disk rather than whatever tempfile provides (again, for large files). Again, this probably won't get used much, but it was more there in case anyone needs it.

Cheers,

Tom

> 
> -- 
> Pauli Virtanen
> 
> _______________________________________________
> SciPy-Dev mailing list
> SciPy-Dev@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev



More information about the SciPy-Dev mailing list