Manticorr is a simple module to calculate the matter correlation function from Gadget outputs. In general, the raw particle data from an N-body simulation is very large. So large in fact that even if a speedy correlation function calculator is available, calculating a matter correlation function from all points is non-trivial. A imple workaround is to sample particles at random from the simulation and calculate the correlation function. Downsampling in this way, though a bit of a hack, allows the entire calculation to easily fit into memory and be done realtively quickly.
Manticorr loads a gadget snapshot, downsamples it, and calculates a correlation function via Corrfunc.
Manticorr has minimal dependencies. It requires numpy (which you probably have) and CorrFunc (which you probably don't). Corrfunc is vailable here.
Manticorr is run via the following command line call:
python manticorr.py [-h] [-p SAMPLE_FRAC] [-n NUM_CORES] inbase outputfile
The parameters are defined via:
inbase: The path and base file name for the gadget files. Manticorr reads all files that satisfyinbase + '.*'.outputfile: The file to save the calculated correlation function to.SAMPLE_FRAC: Optional. Fraction of particles to retain for calculation. Too large of a value will possible exceed memory, and too small will result in too few or 0 particles being sampled!NUM_CORES: Optional. Number of cores to use for the correlation function calculation. Default is the number given by Python'smultiprocessing.num_coresfunction.
Corrfunc requires as an input binfile, a file which contains the lower and upper edges of each bin. Manticorr includes the arbitray default written from np.logspace(-1, 1.7, 20). To change it, simply run makeBins via:
python makeBins.py [-h] rmin rmax nbins
Which plugs those values into their appropriate place in np.logspace and writes them to binfile.