Two-point correlation function

[Khary Richardson]

Hello,

 I have what maybe a trivial question. Can one use the astroML two-point correlation function for data that is in three dimensions?  The reason I ask is that the sample codes all comprise of 2D arrays. 

[Jake Vanderplas]

Hi,

Yes, you can use data of any dimension. For example, the astroML.correlation.two_point() function takes as its first argument a matrix of size [n_samples, n_features]. So, if you have 1000 points in 3 dimensions, you can arrange them in a 1000x3 matrix and pass it to the algorithm,

    Jake

On Wed, Sep 30, 2015 at 2:52 PM, Khary Richardson <surfc...@gmail.com> wrote:

Hello,

 I have what maybe a trivial question. Can one use the astroML two-point correlation function for data that is in three dimensions?  The reason I ask is that the sample codes all comprise of 2D arrays. 

--
You received this message because you are subscribed to the Google Groups "astroML-general" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astroml-gener...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Khary Richardson]

Hi Jake,

   Thank you so much!!

Khary

[Tom Evslin]

if objects are in 3-d space with ra, dec, and z, do coordinates have to be converted to some cartesian form in order retrieve the correlation function values for the bins? If so, if unit are comoving Mpc, will bins than be by Mpc of separation?

On Wednesday, September 30, 2015 at 6:01:33 PM UTC-4, Jake Vanderplas wrote: