scikits.timeseries: saving to binary file

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

scikits.timeseries: saving to binary file

Timmie
Administrator
Hello,
according to the timeseries docs I can save timeseries objects to
a) text files
http://pytseries.sourceforge.net/core.timeseries.io.html#saving-a-timeseries-to-a-text-file
b) hdf5 using pytables
http://pytseries.sourceforge.net/core.timeseries.io.html#saving-a-timeseries-to-a-text-file

Why can I not store it as a numpy binary file using numpy.save without loosing
the dates and frequency information?

Thanks,
Timmie


_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user
Reply | Threaded
Open this post in threaded view
|

Re: scikits.timeseries: saving to binary file

Pierre GM-2

On Sep 2, 2009, at 7:58 AM, Tim Michelsen wrote:

> Hello,
> according to the timeseries docs I can save timeseries objects to
> a) text files
> http://pytseries.sourceforge.net/core.timeseries.io.html#saving-a-timeseries-to-a-text-file
> b) hdf5 using pytables
> http://pytseries.sourceforge.net/core.timeseries.io.html#saving-a-timeseries-to-a-text-file
>
> Why can I not store it as a numpy binary file using numpy.save  
> without loosing
> the dates and frequency information?

Because neither Matt nor I had any need for it so far. We'd be happy  
to consider a patch, of course. Note that you could try to convert the  
series to a structured array (with a 'dates', a 'data' and a 'mask'  
fields) with the "toflex" method and save the resulting ndarray.  
However, you'll probably lose the frequency information (unless you  
find a trick to save meta-information, but as I don't have any  
experience with np.save, I won't be able to help you).
Keep us posted.
P.
_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user