More functions

Functions of class meps.eps

The main functionality of the package was already discussed in previous sections. Here we will learn about 5 additional useful methods of the class meps.eps. These are

  1. The 2-point angular correlation function (2PACF, \(C=C(\chi)\)), acf()

  2. The flux density distribution function (\(\mathrm{d}n/\mathrm{d}S\)), dndS(). As this function is a method of class meps.eps, the choice of the form of \(\mathrm{d}n/\mathrm{d}S\) will have been set when you initialise your class meps.eps object as

from epspy import meps

obj = meps.eps()

  1. Number of source per pixel ( \(n_{\mathrm{ps}}\) ), num_den(). This function has direct correspondence with the form of \(\mathrm{d}n/\mathrm{d}S\) and the choice of \(S_{\mathrm{min}}\) and \(S_{\mathrm{max}}\).

  2. print_input() can be useful if you want see what parameters you are currently running with.

Note that these functions are computationally cheap and thus can be run directly in a jupyter notebook or interactively on the terminal.

Examples

Suppose you want to find the number of sources between \(S=10^{-6}\) and \(10^{-1}\mathrm{Jy}\) for a \(7^{\mathrm{th}}\) order log-log polynomial fit to Euclidean number count then do

>>> from epspy import meps
>>> obj = meps.eps(dndS_form=1, logSmin=-6,logSmax=-1)
>>> Nps = round(obj.Nps)
>>> Nps
249825820

Let us calculate the number density.

>>> nps = obj.num_den()
Total number of sources, Nps = 249825820
Total number of pixels, Npix = 49152
Average number of sources per pixel, n_bar = 5082.72
Done.
Average overdensity for the clustered sky (should be ~ 0) = 0.007
The clustered number density has been saved into file:
 n_ps

(Note that since this is a random process you might get different numbers.)

Let us also check if the sum of the elements of array nps is Nps

>>> ncl.sum()
251605169.7207427

There some discrepancy but do not be worried as this is just a numerical artefact. In fact the fractional error is already printed and is about 0.7%, which is sufficiently small.

Functions of module meps

In case you forgot what data set you generated with what parameter specifications, you can always save your class object using the function save_eps() in the directory where all other outputs are saved and load back using load_eps(). (Both functions are part of module meps.py.)

Thus, after initialising your class object (i.e. obj = meps.eps([YOUR SPECIFICATIONS])), you can add to your script meps.save_eps(obj,'myobj').

Examples

from epspy import meps

obj = meps.eps()
meps.save_eps(obj,'myobj')

Now check if there is a file called myobj.pkl in obj.path directory.

When you came back next time you can load your class object as

from epspy import meps
obj = meps.load_eps('/give/full/path/to/myobj.pkl')

Remember to give the full path to the myobj with the extension .pkl.

Check that indeed the specifications are correctly loaded by printing them using the function print_input().

from epspy import meps
obj = meps.load_eps('/give/full/path/to/myobj.pkl')
obj.print_input()

There is also an argument lbl, which you can use to put an extra label to you output files. For example,

from epspy import meps

obj = meps.eps(lbl='_mylabel')

Now all files names will have _mylabel appended to them. For example, when you run num_den(), the output file name will be called n_ps_mylabel.npy.