[AstroPy] time-series tools in Python

Brigitta Sipocz sic at elte.hu
Sat Mar 5 09:53:41 EST 2016

Hi Hang,

The "Python users in astronomy" facebook group [0] may be a better place to
ask these kind of general questions as there may be more people reading and
answering it there.

The answer may depend on what kind of periodic signals you are looking for.
I have more experience with the following two options below, but there are
several more approaches, too e.g. using wavelets, or gaussian process, etc.

One of the fastest/simplest method would be to run Lomb-Scargle [1], but it
has some serious limitations if your signal is far from being sinusoidal,
or you are going after quasi periodicity with changing signal shape, etc.
There are quite a few python implementation, e.g in scipy or gatspy.
Another method to search for a period is to use the autocorrelation
function. Since your data is unevenly sampled you should use a *discrete
correlation function *instead. Python implementations can be found in the
time_series module of the astroML package.
 Chapter 10 in the book of Statistics, Data Mining, and Machine Learning in
Astronomy by Ivezić et al, is also an excellent guide on the subject.

Hope it helped.


[0] - https://www.facebook.com/groups/astropython/
[1] - https://jakevdp.github.io/blog/2015/06/13/lomb-scargle-in-python/

On 5 March 2016 at 14:20, gonghang.naoc <ghang.naoc at gmail.com> wrote:

> Hi all,
> This may not be a good question posted here, but possibly there should be
> pythoners who work in this field.
> I wonder anybody knows an algorithm for the detection of periodic signals
> in gapped data?
> I searched ADS, but it seems most of the data people deal with are evenly
> sampled.
> Thank you.
> Regards,
> Hang
> _______________________________________________
> AstroPy mailing list
> AstroPy at scipy.org
> https://mail.scipy.org/mailman/listinfo/astropy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/astropy/attachments/20160305/ff5d4802/attachment.html>

More information about the AstroPy mailing list