API Reference

Main Routines (projected space)

These are the main routines for counting pairs and estimating correlation functions. They encapsulate all required steps to provide a true end-user experience : you only need to provide data+parameters to receive final results, and perhaps that amazing plot of your next paper

pcf(tab, tab1, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given two astropy tables corresponding to data and random samples, calculate the two-point projected auto-correlation function (pcf)

All input parameters that control binning, cosmology, estimator, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (pcf) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for DD/RR/DR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.rr, counts.wrp, etc. It also stores the complete log and all input parameters. See Output Dictionary (pcf) for a detailed description.

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data
rans  = Table.read('redgal_rand.fits')                  # read randoms
par = gun.packpars(kind='pcf',outfn='redgal')           # generate default parameters
cnt = gun.pcf(gals, rans, par, write=True, plot=True)   # get pcf and plot
pccf(tab, tab1, tab2, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given three astropy tables corresponding to data, random, and cross samples, this routine calculates the two-point projected cross-correlation function (pccf)

All input parameters that control binning, cosmology, estimator, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

tab2astropy table

Table with cross sample particles (C)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (pccf) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for CD/CR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.cd, counts.cr, counts.wrp, etc. It also stores the complete log and all input parameters. See Output Dictionary (pccf) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data files
rans  = Table.read('redgal_rand.fits')
qsos  = Table.read('qso.fits')
par = gun.packpars(kind='pccf',outfn='redgal_qso')      # generate default parameters
cnt = gun.pccf(gals, rans, qsos, par, write=True )      # get pcf
rppi_A(tab, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given an astropy data table, count pairs in the projected-radial space (\(r_p\) vs \(\pi\) space).

All input parameters that control binning, cosmology, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles

parMunch dictionary

Input parameters. See Input Parameters Dictionary (rppiA) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=True

  • True : generate a plot of counts (integrated radially) vs proj. radius

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.bdd, etc. It also stores the complete log and all input parameters. See Output Dictionary (rppiA) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                     # read data
par = gun.packpars(kind='rppiA',outfn='redgalpairs')  # generate default parameters
cnt = gun.rppiA(gals, par, write=True, plot=True)     # get pair counts and plot
rppi_C(tab, tab1, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given two astropy data tables, cross-count pairs in the projected-radial space (\(r_p\) vs \(\pi\) space).

All input parameters that control binning, cosmology, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with particles (sample D)

tab1astropy table

Table with particles (sample R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (rppiC) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=True

  • True : generate a plot of counts (integrated radially) vs proj. radius

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dr, counts.bdr, etc. It also stores the complete log and all input parameters. See Output Dictionary (rppiC) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
qsos = Table.read('qso.fits')                          # read data
gals = Table.read('redgal.fits')                       # read data
par = gun.packpars(kind='rppiC',outfn='qso_rg_pairs')  # generate default parameters
cnt = gun.rppiC(qsos, gals, par, write=True)           # get pair counts

Main Routines (redshift space)

These are the main routines for counting pairs and estimating correlation functions. They encapsulate all required steps to provide a true end-user experience : you only need to provide data+parameters to receive final results, and even that amazing plot of your next paper

rcf(tab, tab1, par, nthreads=-1, write=True, para=False, plot=False, **kwargs)[source]

Given two astropy tables corresponding to data and random samples, this routine calculates the two-point redshift space auto-correlation function (rcf)

All input parameters that control binning, cosmology, estimator, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (rcf) for a detailed description

writebool. Default=True

  • True : write several output files for DD/RR/DR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

parabool. Default=False

  • True : run in parallel over all available ipyparallel engines

  • False : run serially. No need for any ipyparallel cluster running

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.rr, counts.xis, etc. It also stores the complete log and all input parameters. See Output Dictionary (rcf) for a detailed description.

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data
rans  = Table.read('redgal_rand.fits')                  # read randoms
par = gun.packpars(kind='rcf',outfn='redgal')           # generate default parameters
cnt = gun.rcf(gals, rans, par, write=True, plot=True)   # get rcf and plot
rccf(tab, tab1, tab2, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given three astropy tables corresponding to data, random, and cross samples, this routine calculates the two-point redshift space cross-correlation function (rccf)

All input parameters that control binning, cosmology, estimator, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

tab2astropy table

Table with cross sample particles (C)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (rccf) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for CD/CR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.cd, counts.cr, counts.xis, etc. It also stores the complete log and all input parameters. See Output Dictionary (rccf) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data files
rans  = Table.read('redgal_rand.fits')
qsos  = Table.read('qso.fits')
par = gun.packpars(kind='pccf',outfn='redgal_qso')      # generate default parameters
cnt = gun.rccf(gals, rans, qsos, par, write=True )      # get rcf
s_A(tab, par, nthreads=-1, write=True, para=False, plot=False, **kwargs)[source]

Given an astropy table, count pairs in redshift space

All input parameters that control binning, cosmology, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles

parMunch dictionary

Input parameters. See Input Parameters Dictionary (sA) for a detailed description

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

parabool. Default=False

  • True : run in parallel over all available ipyparallel engines

  • False : run serially. No need for any ipyparallel cluster running

plotbool. Default=False

  • True : generate a plot of counts vs redshift separation

  • False : do not generate any plots

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.bdd, etc. It also stores the complete log and all input parameters. See Output Dictionary (sA) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                   # read data
par = gun.packpars(kind='sA',outfn='redgalpairs')   # generate default parameters
cnt = gun.s_A(gals, par, write=True, plot=True  )   # get counts and plot
s_C(tab, tab1, par, nthreads=-1, write=True, para=False, plot=False, **kwargs)[source]

Given two astropy tables, cross-count pairs in redshift space

All input parameters that control binning, cosmology, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with particles (sample D)

tab1astropy table

Table with particles (sample R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (sC) for a detailed description

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

parabool. Default=False

  • True : run in parallel over all available ipyparallel engines

  • False : run serially. No need for any ipyparallel cluster running

plotbool. Default=False

  • True : generate a plot of counts vs redshift separation

  • False : do not generate any plots

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dr, counts.bdr, etc. It also stores the complete log and all input parameters. See Output Dictionary (sC) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
qsos = Table.read('qso.fits')                        # read data
gals = Table.read('redgal.fits')                     # read data
par = gun.packpars(kind='sC',outfn='qso_rg_pairs')   # generate default parameters
cnt = gun.rppiC(qsos, gals, par, write=True)         # get pair counts

Main Routines (angular space)

These are the main routines for counting pairs and estimating correlation functions. They encapsulate all required steps to provide a true end-user experience : you only need to provide data+parameters to receive final results, and even that amazing plot of your next paper

acf(tab, tab1, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given two astropy tables corresponding to data and random samples, this routine calculates the two-point angular space auto-correlation function (acf)

All input parameters that control binning, cosmology, estimator, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (acf) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for DD/RR/DR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.rr, counts.xis, etc. It also stores the complete log and all input parameters. See Output Dictionary (acf) for a detailed description.

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data
rans  = Table.read('redgal_rand.fits')                  # read randoms
par = gun.packpars(kind='rcf',outfn='redgal')           # generate default parameters
cnt = gun.rcf(gals, rans, par, write=True, plot=True)   # get rcf and plot
accf(tab, tab1, tab2, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given three astropy tables corresponding to data, random, and cross samples, this routine calculates the two-point angular space cross-correlation function (accf)

All input parameters that control binning, estimator, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles (D)

tab1astropy table

Table with random particles (R)

tab2astropy table

Table with cross sample particles (C)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (accf) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for CD/CR counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate the CF plot on screen (and saved to disk when write=True)

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.cd, counts.cr, counts.wth, etc. It also stores the complete log and all input parameters. See Output Dictionary (accf) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                       # read data files
rans  = Table.read('redgal_rand.fits')
qsos  = Table.read('qso.fits')
par = gun.packpars(kind='accf',outfn='redgal_qso')      # generate default parameters
cnt = gun.accf(gals, rans, qsos, par, write=True )      # get accf
th_A(tab, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given an astropy data table, count pairs in angular space

All input parameters that control binning, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with data particles

parMunch dictionary

Input parameters. See Input Parameters Dictionary (thA) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate a plot of counts vs angular separation

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dd, counts.bdd, etc. It also stores the complete log and all input parameters. See Output Dictionary (thA) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
gals  = Table.read('redgal.fits')                   # read data
par = gun.packpars(kind='thA',outfn='redgalpairs')  # generate default parameters
cnt = gun.th_A(gals, par, write=True, plot=True  )  # get counts and plot
th_C(tab, tab1, par, nthreads=-1, write=True, plot=False, **kwargs)[source]

Given two astropy data tables, cross-count pairs in angular space

All input parameters that control binning, column names, etc. are passed in a single dictionary par, which can be easily generated with default values using gundam.packpars() and customized later

Parameters

tabastropy table

Table with particles (sample D)

tab1astropy table

Table with particles (sample R)

parMunch dictionary

Input parameters. See Input Parameters Dictionary (thC) for a detailed description

nthreadsinteger. Default=-1

Number of threads to use. Default to -1, which means all available cpu cores

writebool. Default=True

  • True : write several output files for counts, correlations, etc.

  • False : do not write any output files (except for logs, which are always saved)

plotbool. Default=False

  • True : generate a plot of counts vs angular separation

  • False : do not generate any plots

kwargsdict

Extra keywords passed to gundam.plotcf()

Returns

countsMunch dictionary

Munch dictionary, containing all counts and correlations accesible by field keys, e.g. counts.dr, counts.bdr, etc. It also stores the complete log and all input parameters. See Output Dictionary (thC) for a detailed description

Examples

import gundam as gun ; from astropy.table import Table
qsos = Table.read('qso.fits')                        # read data
gals = Table.read('redgal.fits')                     # read data
par = gun.packpars(kind='thC',outfn='qso_rg_pairs')  # generate default parameters
cnt = gun.th_C(qsos, gals, par, write=True)          # get pair counts

Auxiliary Routines

These are helper routines that make it easy to work with correlations, visualize count data, optimize grid size, sort data, etc. For examples of how to use them check their individual help, or even better, the source code of gundam.pcf()

bestSKgrid2d(par, npts, ras, dens=None)[source]

Try to find the optimum size (mxh1,mxh2) of a 2D skip grid, for an arbitrary sample of (ra,dec) points.

This is far from trivial, so for the moment, this routine works as follows:

  1. Choose an “optimum” target cell density (around 22 ppcell in many tests)

  2. Find the best mxh1 from an empirical best fit relation of npts vs mxh1

  3. Adjust mxh2 to reach the target cell density

Parameters

parMunch dictionary

Input parameters dictionary

nptsinteger or list of integers

Nr. of objects in the sample(s). For cross-correlations it should be a list of the form [sample_D_size, sample_R,size] and the effective npts adopted is their sum

rasfloat array or list of arrays

Right ascention of objects in the samples(s). For cross-correlations, the two samples are concatenated

densfloat. Default=None

Target cell density. Note that dens=None implicates a default value of 22 if npts>50000 and 16 otherwise.

Returns

mxh1, mxh2integer

Nr. of DEC and RA cells of the SK table, respectively

densfloat

The effective cell density adopted

bestSKgrid3d(par, npts, ras, dens=None)[source]

Try to find the optimum size (mxh1,mxh2,mxh3) of a 3D skip grid, for an arbitrary sample of (ra,dec,dcom) points.

This is far from trivial, so for the moment, this routine works as follows:

  1. Choose an “optimum” target cell density according to the type of correlation

  2. Choose the best mxh3 as mxh3=int((dcmax-dcmin)/rvmax)

  3. Find the best mxh1 from an empirical best fit relation of npts vs mxh1

  4. Adjust mxh2 to reach the target cell density

Parameters

parMunch dictionary

Input parameters dictionary

nptsinteger or list of integers

Nr. of objects in the sample(s). For cross-correlations it should be a list of the form [sample_D_size, sample_R,size] and the effective npts adopted is their sum

rasfloat array or list of arrays

Right ascention of objects in the samples(s). For cross-correlations, the two samples are joined together

densfloat. Default=None

Target cell density. Note that dens=None implicates different default values. See function code.

Returns

mxh1, mxh2, mxh3integer

Nr. of DEC, RA, DCOM cells of the SK table, respectively

densfloat

The effective cell density adopted

cfindex(path='./')[source]

List file_name + description of all count (.cnt) files in a given path. Useful to quickly check a dir with dozens of correlation function runs and therefore hundreds of files.

Parameters

pathstring

Path to list descriptions of .cnt files inside

cnttable(cfi, fmt1=None, fmt2=None, write=False, browser=True)[source]

Shows a nicely formatted tabular view of the count data stored in a counts output dictionary. The table is printed in stdout and optionally displayed in the default web browser.

Parameters

cfistring or Munch dictionary

Filepath for the counts (.cnt) file, or the counts dictionary itself

fmt1string

Ouput format of numeric fields (bins, correlations and errors). Default=’.4f’

fmt2string

Ouput format of numeric fields (counts). Default=’.2f’

writebool

If write=True, save table to disk. Filename will be asked for. Default=False

browserbool

If browser=True, display HTML table in browser. Default=True

Returns

tabastropy table

Single table with all relevant counts as columns. Use print(tab) or tab.show_in_browser()

Examples

# show info for a projected correlation from file on disk
cnttable('/proj/galred.cnt')
# show info a from variable in the session's memory
cnttable(galred)
makebins(nsep, sepmin, dsep, logsep)[source]

Create arrays of bins in which Gundam will count pairs and estimate correlation functions, given the number of bins desired, the minimum bin value and the chosen bin width.

Note units are not needed, but should be interpreted according to the input parameters

Parameters

nsepinteger

Number of bins

sepminfloat

Minimum bin location

dsepfloat

Bin width (dex if logsep=1)

logsepbool

If True, do log-space binning. If False, do linear-space binning

Returns

sepfloat array

Bin locations used by Gundam (left-side + extra bin at right limit)

sepoutfloat array

Left, middle and right-side points of each bin. Useful to plot more easily

Examples

# Create 18 log bins of size=0.2 dex in redshift and projected space
seps,sepsout = makebins(18,0.01,0.2,1)
sepp,seppout = makebins(18,0.01,0.2,1)
# Create 25 bins of size 2 Mpc in radial space, from 0 to 50 Mpc
sepv = makebins(25,0.,2.,0)[0]
# Create instead 1 bin of size 50 Mpc, e.g. to work out the pcf integrated from 0 to 50 Mpc
sepv = makebins(1,0.,50.,0)[0]
pixsort(tab, colnms, par)[source]

Arrange an astropy table with (ra,dec) or (ra,dec,z) data into a grid of SK pixels and sort the rows of the table according to the pixel index, ordered by a given methodology. This way data in a given pixel sits closer in memory, largely increasing the efficiency of the cache.

Several (experimental) orders such as Morton and Hilbert are available

Parameters

tab: astropy table

Data table with ra/dec or ra/dec/z points

colnms: list of strings

Colum names for coordinates and redshift, e.g. [‘ra’,’dec’,’z’], [‘RAJ2000’,’DEJ2000’,’redshift’]

par: Munch dictionary

Input parameters dictionary. Must contain mxh1, mxh2, mxh3 (when appropiate), the survey boundaries sbound and the desired method for ordering pxorder. For samples that cross the RA=0 limit, it can is useful to specify a custom RA boundary. See Custom RA boundaries

Returns

sidxarray

Sort index. Use tab=tab[sidx] to actually sort the data

qprint(self)[source]

Prints a quick, nicely formatted version of Munch dictionaries, such as the counts ouput dictionary or the par input dictionary used by Gundam routines. Very useful to quickly explore what is stored inside.

This routine is added dynamically to the Munch class, so it can be accessed as any_munch_obj.qprint()

Note counts output dictionaries can also be explored using gundam.cnttable() to display a more elaborated tabular view of (exclusively) the counts calculated.

Examples

import gundam as gun

# Explore the many arrays and counts of a typical run
cred = gun.readcounts('redgalaxies.cnt')
cred.qprint()

# Check the parameters used to create the run
cred.par.qprint()
radec2xyz(ra, dec, r=0.5)[source]

Converts a (ra,dec) coordinates to rectangular (x,y,z) coordinates in a sphere of radius r.

For r=0.5, this allows to speed up subsecuent haversine distance calculations between two points, by simply computing \(dhav^2 = (x1-x2)^2 + (y1-y2)^2 + (z1-z2)^2\)

Parameters

ra,decfloat arrays

Right ascention an declination coordinates

rfloat. Default=0.5

Sphere radius

Returns

(x,y,z)tuple

Rectangular coordinates

tpcf(npt, nrpt, dd, bdd, rr, dr, estimator)[source]

Return the (auto)correlation function for a given estimator and arrays of data and random counts

If DR counts are not needed (e.g. the ‘NAT’ estimator), just set dr=0

If boostrap errors are not needed or available, just set bdd to a zero-valued array with null 2nd dimension, e.g. bdd=np.zeros([len(dd),0])

Parameters

npt,nrptinteger

Number of data and random particles

ddfloat array

DD counts

bddfloat array

Bootstrap DD counts

rrfloat array

RR counts in projected and radial separations

drfloat array

DR counts in projected and radial separations

estimatorstring

Statistical estimator of the correlation function. Default=`NAT`

  • ‘NAT’ : Natural -> \(DD/RR - 1\)

  • ‘HAM’ : Hamilton -> \(DD*RR/DR^{2} - 1\)

  • ‘LS’ : Landy-Szalay -> \((DD - 2DR + RR) / RR\)

  • ‘DP’ : Davis-Peebles -> \(DD/DR - 1\)

Returns

xifloat array

Correlation function

xierrfloat array

Boostrap error estimate. Set to zero if bdd is nulled as explained above

Notes

See this paper for a nice review on estimators and their normalization factors. Here, the normalization factors are derived to : (1) keep estimator formulae clean, (2) avoid having operations such as (npt*(npt-1)) * dd, where counts are multiplied/divided by very big numbers when npt is large.

Examples

# Calculate the angular CF using the Landy-Szalay estimator
acf, acferr = gun.tpcf(npt,nrpt,dd,bdd,rr,dr,estimator='LS')
tpccf(npt, nrpt, cd, bcd, cr, estimator)[source]

Return the (cross)correlation function for a given estimator and count arrays for data (D), random (R) and cross (C) samples.

For the moment the only estimator implemented is the Davis-Peebles : \(\xi=CD/CR-1\)

If bootstrap errors are not needed or available, just set bdd to a zero-valued array, e.g. bdd=np.zeros([len(dd),0])

Parameters

npt,nrptinteger

Number of particles in data (D) and random (R) samples

cdfloat array

CD counts

bcdfloat array

Bootstrap CD counts

crfloat array

CR counts

estimatorstring

  • ‘DP’ : Davis-Peebles -> \(CD/CR-1\)

Notes

C and D are data samples while R is random sample corresponding to D

Returns

fxifloat array

Cross-correlation function

fxierrfloat array

Boostrap error estimates

Examples

import gundam as gun
c = gun.readcounts('qso_gal.cnt')

(ccf,ccferr) = tpccf(c.npt, c.nrpt, c.cd, c.bcd, c.cr, estimator='DP')
tpcf_wrp(npt, nrpt, dd, bdd, rr, dr, dsepv, estimator)[source]

Return the projected (auto)correlation function for a given estimator and arrays of data and random counts

If DR counts are not needed (e.g. the ‘NAT’ estimator), just set dr=0

If boostrap errors are not needed or available, just set bdd to a zero-valued array with null 2nd dimension, e.g. bdd=np.zeros([len(dd),0])

Parameters

npt,nrptinteger

Number of data and random particles

ddfloat array

DD counts in projected and radial separations

bddfloat array

Bootstrap DD counts in projected and radial separations

rrfloat array

RR counts in projected and radial separations

drfloat array

DR counts in projected and radial separations

dsepvfloat

Bin size in radial direction

estimatorstring

Statistical estimator of the correlation function

  • ‘NAT’ : Natural -> \(DD/RR - 1\)

  • ‘HAM’ : Hamilton -> \(DD*RR/DR^{2} - 1\)

  • ‘LS’ : Landy-Szalay -> \((DD - 2DR + RR) / RR\)

  • ‘DP’ : Davis-Peebles -> \(DD/DR - 1\)

Returns

wrpfloat array

Correlation function

wrperrfloat array

Boostrap error estimate. Set to zero if bdd is nulled as explained above

Notes

See this paper for a nice review on estimators and their normalization factors. Here, the normalization factors are derived to : (1) keep estimator formulae clean, (2) avoid having operations such as (npt*(npt-1)) * dd, where counts are multiplied/divided by very big numbers when npt is large.

Examples (xxxx TODO)

(wrp,wrperr) = tpcf_wrp(npt,nrpt,ddpv,bddpv,rrpv,drpv,dsepv,estimator='HAM')
tpccf_wrp(npt, nrpt, cd, bcd, cr, dsepv, estimator)[source]

Return the projected (cross)correlation function for a given estimator and count arrays

If boostrap errors are not needed or available, just set bcd to a zero-valued array with null 2nd dimension, e.g. bcd=np.zeros([len(ddXXX),0])

Parameters

nptinteger

Number of data particles

nrptinteger

Number of random particles

cdfloat array

CD counts in projected and radial separations

bcdfloat array

Bootstrap CD counts in projected and radial separations

crfloat array

CR counts in projected and radial separations

dsepvfloat

Radial bin size

estimatorstring

Statistical estimator of the correlation function

  • ‘DP’ : Davis-Peebles -> \(CD/CR - 1\) (C,D data samples, R is random of D)

Returns

wrpfloat array

Projected cross-correlation function

wrperrfloat array

Boostrap error estimate

Examples

# remains to do XXXXX
(wrp,wrperr) = tpccf_wrp(npt,nrpt,cd,bcd,cr,dsepv,estimator='DP')

Plotting Routines

cntplot(cnt, **kwargs)[source]

Plot a correlation function from a counts output dictionary (either read from disk or passed directly). Both axes are set to log-space and axes labels are selected automatically according to the type of correlation (i.e. given by par.kind)

This is a wrapper for gundam.plotcf(), so all of its parameters can be specified too.

Parameters

cntstring or Munch dictionary

Filepath for the counts (.cnt) file, or the counts dictionary itself

kwargskeyword list

Any extra [key]=value pairs are passed to the underlying gundam.plotcf() routine

Examples

import gundam as gun

# Read a pcf run and plot the correlation function
cnt1 = gun.readcounts('/p01/redgalsP.cnt')
cntplot(cnt1)
# Plot the correlation function from a .cnt file
cntplot('/p01/redgalsA.cnt', label='angcf of redgals', fill=True)
cntplot2D(cnt, estimator=None, slevel=5, write=False, figfile=None, xlabel=None, ylabel=None, cmap='jet', **kwargs)[source]

Plot the 2D correlation function in the projected-radial space (\(r_p\) vs \(\pi\) space) with optional gaussian smoothing and contour levels

Parameters

cntstring or Munch dictionary

Filepath for the counts (.cnt) file or the counts output dictionary

estimatorstring. Default=None

Estimator for the correlation function. Any of (‘NAT’,’LS’,’HAM’,’DP’). If estimator=None, then it is taken from cnt.par.estimator

slevelfloat. Default=5

Smoothing level (namely the size of the Gaussian smothing kernel)

writebool. Default=False

Save the figure to disk (default format is pdf). See Notes to save in other graphic formats

figfilestring. Default=None

Specify an alternative file name for the figure. If None, then choose cnt.par.outfn as default. Do not add extension.

xlabel, ylabelstring. Default=None

X-axis and Y-axis labels. If supplied, they override the default labels (\(r_p \ [h^{-1} Mpc]\) and \(\pi \ [h^{-1} Mpc]\))

cmapstring. Default=’jet’

Colormap for the plot

kwargskeyword list

Any extra [key]=value pairs are passed to matplolib.pyplot.pcolor() Use this to customize shading, edges, alpha, etc.

Notes

  • The graphic format can be changed by passing the figformat key in kwargs, e.g. figformat='pdf'. Any format supported by matplotlib is valid.

Examples

# Check some nice Fingers of God and the Kaiser squashing
cntplot2D('lum_red_gals.cnt', cmap='viridis')
comparecf(clist1, clist2=None, shift=0.0, fac=1.0, ploterrbar=True, fill=False, filtneg=False, label=None, plotratio=False, ratioxrange=None, color1=None, marker1=None, markers1=None, linestyle1=None, linewidth1=None, color2=None, marker2=None, markers2=None, linestyle2=None, linewidth2=None, f=None, ax1=None, ax2=None)[source]

Plot multiple correlation functions in a single figure for easy comparison. Optionally show an additional lower panel displaying the ratio of each function respect to a single “control” correlation (many-to-one) or to multiple correlations (one-to-one).

Parameters

clist1list of Munch dictionaries / list of strings

Count dictionaries or filepaths of .cnt files of correlations

clist2list of Munch dictionaries / list of strings. Default=None

List of control correlations. When plotratio=True, the y-values of each correlation curve in clist1 are divided by those in clist2 (one-to-one) and plotted in a lower panel. If clist2 has a single element, the ratios are from all clist1 divided by the single clist1. See Notes for more details

shiftfloat. Default=0.0

Fraction of bin size by which x values are shifted. Useful to slightly separate overlapping curves

facfloat. Default=1.0

Multiplication factor for y and yerr

ploterrbarbool. Default=True

If ploterrbar=True, plot error bars according to yerr

fillbool. Default=False

If fill=True, plot a filled semi-transparent error region instead of the usual error bars

filtnegbool. Default=False

If filtneg=True, filter out points where (y-yerr)<0, i.e. those with large errors in a log plot

label: list

Optional list of strings to label each correlation function. If ommited, the values are taken from the outfn key stored in the count objects

plotratiobool. Default=False

If plotratio=True, plot also a lower panel with ratios of clist1 respect to clist2

ratioxrangelist of form [xmin,xmax]

Only plot the ratio between xmin and xmax

color1,marker1,markers1,linestyle1,linewidth1lists

List of colors, marker symbols, marker sizes, line styles and line widths for curves in clist1

color2,marker2,markers2,linestyle2,linewidth2lists

List of colors, marker symbols, marker sizes, line styles and line widths for control curves in clist2

ffigure instance

Handle of existing Figure instance

ax1,ax2axes instances

Handles of correlation plot and ratio plot axes

Notes

The correlation curves in clist2 are not plotted in the correlation function panel while curves present in both clists are not shown in the ratio panel (i.e. to avoid ratios of curves respect to themselves).

Returns

(f,ax1,ax2) or (f,ax1)tuple of handles

Handles of figure, correlation axis (ax1) and ratio axis (ax2), if present

Examples

# Compare two w(rp) correlations
comparecf(['galred.cnt', 'galblue.cnt'])
# Compare one acf on disk with another passed as a counts ouput dictionary
comparecf(['/proj/galred.cnt', qso])
# Compare multiple samples and plot sample/control ratios
f,ax1,ax2 = comparecf(['galred.cnt', 'galblue.cnt', 'galgreen.cnt'], clist2=['allgals.cnt'], fill=True, plotratio=True)
# Add another curve to previous plot
comparecf(['qso.cnt'], clist2=['control_qso.cnt'], color2=['k'], f=f, ax1=ax1, ax2=ax2, plotratio=True)
fitpowerlaw(x, y, yerr, iguess=[1.0, -1.0], fitrange=None, plot=False, markfitrange=False, **kwargs)[source]

Fit a power-law of the form \(ax^{\gamma}\) to a correlation function over a given x-coordinate range. Optionally plot the fitted curve

Parameters

xfloat array

x-coordinates such as cnt.rpm, cnt.thm, etc.

yfloat array

x-coordinates such as cnt.wrp, cnt.wth, etc.

yerrfloat array

Errors in y-coordinates such as cnt.wrperr, cnt.wtherr, etc.

iguesslist of floats. Default=[1., -1.]

Initial guesses for \(a\) and \(\gamma\)

fitrangefloat array of form [xmin,xmax]. Default=None

Restrict fit to points inside the given interval

plotbool. Default=False

Plot the fitted power-law curve

markfitrangebool. Default=False

Overlay marks for points actually used in the fitting

kwargskeyword list

Any extra [key]=value pairs are passed to matplolib.pyplot.plot() Use this to customize colors, linestyles, markers, etc.

Examples

import gundam as gun

c1 = gun.readcounts('galaxies.cnt')
cntplot(c1)
gun.fitpowerlaw(c1.rpm, c1.wrp, c1.wrperr, plot=True)
plotcf(x, y, yerr, fac=1.0, write=False, figfile=None, par=None, angunit='deg', xlabel=None, ylabel=None, label=None, shift=0.0, ploterrbar=True, fill=False, filtneg=False, **kwargs)[source]

Plot a correlation function from arrays of x, y and yerr values. Both axes are set to log-space and axes labels are selected automatically according to the type of correlation (i.e. given by par.kind)

Parameters

x,y,yerrfloat arrays

x, y coordinates and corresponding errors of the correlation function. If yerr=0 or all elements of yerr are <=0 or ploterrbar=False, no errorbar is plotted

facfloat. Default=1.0

Multiplication factor for y and yerr

writebool. Default=False

Save the figure to disk (default format is pdf). See Notes to save in other graphic formats

figfilestring. Default=None

Specify an alternative file name for the figure. If specified, overrides the default which is to take it from par.outfn. Do not add extension.

pardictionary of type Munch. Default=None

Used to pass outfn name to name saved figures when write=True

angunitstring. Default=’deg’

  • ‘arcsec’ : set ouput axis in arcsec (x values are unscaled)

  • ‘arcmin’ : set ouput axis in arcmin (x values are scaled as x/60)

  • ‘deg’ : set ouput axis in degrees (x values are scaled as x/3600)

xlabel, ylabelstring. Default=None

X-axis and Y-axis labels. If supplied, they override the default labels deduced from par.kind

labelstring. Default=None

Label for the legend of the curve. If supplied, it will override the default label which is the basename of par.outfn. Note you have to issue at least a plt.legend() to actually display the legend box

shiftfloat. Default=0.0

Fraction of bin size by which x values are shifted. Useful to slightly separate overlapping curves

ploterrbarbool. Default=True

If ploterrbar=True, plot error bars according to yerr

fillbool. Default=False

If fill=True, plot a filled semi-transparent error region instead of the usual error bars

filtnegbool. Default=False

If filtneg=True, filter out points where (y-yerr)<0, i.e. those with large errors in a log plot

kwargskeyword list

Any extra [key]=value pairs are passed to the underlying matplotlib.pyplot.plot() routine, except for alpha which is passed to pyplot.fill_between(), capsize which is passed to pyplot.errorbar(), and figformat which is passed to pyplot.savefig(). Use this to customize colors, linestyles, markers, etc.

Notes

  • Sucessive calls cycle between 4 predefined styles (for color, marker, linewidth, etc.) that can be overrriden by passing the corresponding [key]=value pairs in kwargs

  • The output graphic format can be changed by passing the figformat key in kwargs, e.g. figformat='pdf'. Any format supported by matplotlib is valid.

Examples

import gundam as gun
c1 = gun.readcounts('redgalPCF.cnt')
c2 = gun.readcounts('redgalRCF.cnt')

plt.figure(1)   # Plot w(rp)
gun.plotcf(c1.rpm,c1.wrp,c1.wrperr,par=c1.par)
plt.figure(2)   # Plot xi(s)
gun.plotcf(c2.sm,c2.xis,c2.xiserr,par=c2.par,color='yellow')

Comprehensive Auxiliary Routines

These are helper routines that perform many common tasks (in the Python side) during a correlation run, such as collecting counts, i/o functions, initialization, logging, etc. For a nice example of how to use them, see source code of gundam.pcf()

addlog(file, key, m)[source]

Read a text file and dump it into a key of a given dictionary. Useful to add entire logs from disk into Munch objects

Parameters

filestring

Complete path of file, usually any log file

keystring

Name of the key to be created

mMunch dictionary

The dictionary where the key m.key will be created

allequal(v)[source]

Fast way to check if all elements of a 1D-array have the same value. Useful to detect when all weights are set to 1.0, and hence to call faster versions of the counting routines

Parameters

varray_like

Array to be checked

Returns

resbool

True if all elements have the same value

addpvcols(x, cfo, basecolname, **kwargs)[source]

Auxiliary function used by gundam.cnttable() to append columns to a table populated with the fields of counts ouput dictionary that store pair counts, all with nice column names. Works with 1d counts or 2d counts arrays (e.g. those from a pcf run when nsepv>1).

Parameters

xastropy table

Table to add data

cfoMuch dictionary

Counts dictionary with the count arrays, e.g. cfo.dd, cfo.rr, etc.

basecolnamestring

The name of the field to add, e.g. dd, and also the prefix for the column name, which if needed will be appended with _001, _002, etc. for each radial bin

kwargs :

Any [key]=value pair to pass to the astropy Column constructor. Intended to pass a format specification for the column, such as format='.4f'

Returns

None, it modifies the input table x

bound2d(decs)[source]

Return the (maximal) survey boundaries in (RA,DEC) for multiple samples. Note in the RA direction, the limits are always set as ramin=0. and ramax=360. to make the pair counting valid for surveys that cross the origin of coordinates

Parameters

decsfloat array or list of arrays

DEC coordinates of one or more samples [deg]

Returns

boundtuple

The limits (ramin,ramax,decmin,decmax) that enclose all input samples

bound3d(decs, dcs)[source]

Return the (maximal) survey boundaries in (RA,DEC,DCOM) for multiple samples. Note in the RA direction, the limits are always set as ramin=0. and ramax=360. to make the pair counting valid for surveys that cross the origin of coordinates

Parameters

decsfloat array or list of arrays

DEC coordinates of one or more samples [deg]

dcsfloat array or list of arrays

Comoving distances of one or more samples [Mpc/h]

Returns

boundtuple

The limits (ramin,ramax,decmin,decmax,dcmin,dcmax) that enclose all input samples

cross0guess(ra)[source]

Guess if a set of RA coordinates cross the RA=0 division (by finding one source 1deg to the left and another 1deg to the right of RA=0, at least)

Parameters

raarray

Right ascention coordiantes

Returns

resbool

True if the sample seems to cross the RA=0 boundary

buildoutput(par, npts=[], binslmr=[], dd=None, rr=None, dr=None, cd=None, cr=None, bootc=None, cf=None, cferr=None)[source]

Given a set of pair count arrays, assemble the final counts output dictionary, adding also bins, input parameters and relevant sample sizes.

This is for the main Gundam functions that calculate a multiple pair counts, i.e. gundam.pcf(), gundam.pccf(), gundam.rcf(), gundam.rccf(), gundam.acf(), gundam.accf()

Parameters

parMunch dictionary

Input parameters

nptslist of int. Default=[]

A list with the nr. of points in each sample, i.e. [npt, npt1]. For example npts=[400, 4000] in cross-count cases; npts=[400] for auto-count cases

binslmrlist. Default=[]

The left, mid and right-side locations of each bin, returned of example by gundam.makebins()

dd,rr,drarray. Default=None

The “dd”, “rr” and “dr” count arrays, if available

cd,crarray. Default=None

The “cd” and “cr” count arrays, if available

bootcarray. Default=None

The boostrap count array, if available

cf,cferrarray. Default=None

The correlation function and its error

Returns

countsMunch dictionary

The ouput dictionary

buildoutputC(par, npts=[], binslmr=[], dd=None, dr=None, bootc=None, intpi=None, intpib=None)[source]

Given a set of pair count arrays, assemble the final counts output dictionary, adding also bins, input parameters and relevant sample sizes.

This is for the main Gundam functions that calculate a single pair count, i.e. gundam.rppi_A(), gundam.rppi_C(), gundam.s_A(), gundam.s_C(), gundam.th_A(), gundam.th_C()

Parameters

parMunch dictionary

Input parameters

nptslist of int. Default=[]

A list with the nr. of points in each sample, i.e. [npt, npt1]. For example npts=[400, 4000] in cross-count cases; npts=[400] for auto-count cases

binslmrlist. Default=[]

The left, mid and right-side locations of each bin, returned of example by gundam.makebins()

dd,drarray. Default=None

The “dd” and “dr” count arrays, if available

bootcarray. Default=None

The boostrap count array, if available

intpi, intpibarray. Default=None

“dd” counts and boostrap “dd” counts integrated along all radial bins, if available

Returns

countsMunch dictionary

The ouput dictionary

check_kind(par, kind)[source]

Check that par.kind = kind. Useful to test, for example, if you are passing the right par object before really counting pairs

Parameters

parMunch dictionary

Input parameters dictionary for Gundam routines

kindstring

The kind to check against, e.g. ‘pcf’, ‘accf’, ‘rppiA’, etc.

closelog(log, runspyder=True)[source]

Close the log machinery used by the main counting routines by removing the handlers

Parameters

loglogging object

The log object

runspyderbool. Default=True

If runspyder=True, remove the extra handler for stdout added when running under a Spyder console

finalize(log, logf, logff, Ltime, t0, counts)[source]

Perform some finalization tasks common to all correlation runs, by logging loop/total times and adding the contents of log files to the counts output dictionary

Parameters

loglogging object

The log object

logfilestring

The complete path of the .log file

Ltime, Ttimefloat

Loop time and total compute time of a correlation run

countsMunch dictionary

Output dictionary containing all counts and correlations

initialize(kind, par, nthreads=None, write=None, plot=None)[source]

Perform some initialization tasks common to all correlation function runs, such as reading par from disk file if needed, check the parameter kind, find out if running under Spyder, initialize the logs, etc.

Parameters

kindstring

The kind to check against, e.g. ‘pcf’, ‘accf’, ‘rppiA’, etc.

parMunch dictionary

Input parameters dictionary for Gundam routines

nthreadsinteger.

Number of threads to use. Passed just to store it under par

writebool. Default=True

Flag to generate output count files. Passed just to store it under par

plotbool. Default=True

Flag to generate plot. Passed just to store it under par

Returns

parMunch dictionary

Input parameters dictionary for Gundam routines

loglogging object

The log object

logfile, logfilefortstring

The complete path of the .log file and the .fotran.log file

runspyderbool. Default=True

Detect if running under a Spyder console

logcallinfo(log, par, npts=[])[source]

Write to log some useful runtime parameters of the main counting routines

Parameters

loglogging object

The log object

parMunch dictionary

Input parameters dictionary for Gundam routines

nptslist of 1, 2 or 3 integers

The nr of objects in the input data table, random table, and cross sample table

logtimming(log, cntid, t)[source]

Write a message to a log instance, showing the compute time for the counts identified with a certain ID string

Parameters

loglogging object

Log object

cntidstring

String to ID a certain type of counts, e.g. ‘DD’, ‘RR’, DR’, etc.

tfloat

The time elapsed [s]

readcounts(cfile, silent=False)[source]

Read from disk the counts dictionary generated by the main counting routines.

Parameters

cfilestring

Filepath for the counts (.cnt) file

silentbool

If False, print a status message indicating the file was read. Default=False

Returns

countsMunch dictionary

The counts object

readpars(filename)[source]

Load from a JSON (.par) file the input parameters dictionary used by many Gundam routines.

Parameters

filenamestring

Filepath of .par file

savecounts(cnt, altname=None)[source]

Save to disk the counts output dictionary returned by the main counting routines.

The default file name is cnt.par.outfn + .cnt, which can be overriden as altname + .cnt, if supplied

Parameters

cntMunch dictionary

The counts object

altnamestring. Default=None

If supplied, use an alternative file name instead of cnt.par.outfn

savepars(par, altname=None)[source]

Save the parameters dictionary par, such as the one generated by gundam.packpars(), in a JSON file. By default it is named as par.outfn + .par

Parameters

parMunch dictionary

Input parameters dictionary for Gundam routines

altnamestring. Default=None

If supplied, use an alternative file name instead of par.outfn + .par

Examples

import gundam as gun

# Get default values for an angular CF run and save to disk
par = gun.packpars(kind='acf', outfn='/proj/acfrun01')
gun.savepars(par)
setlogs(par, runspyder=True)[source]

Set up the log machinery used by the main counting routines by creating the logger object, adding the required handlers and cleaning previous logs if present

Parameters

parMunch dictionary

Input parameters dictionary for Gundam routines

runspyderbool. Default=True

If runspyder=True, add an extra handler for stdout when running under a Spyder console

Returns

loglogging object

The log object

logfilestring

The complete path of the .log file

writeasc_cf(lb, mb, rb, f, ferr, par, fmt='%17.5f', altname=None)[source]

Write an ASCII file for w(rp) / xi(s) / w(th) ouput counts produced by the code

Parameters

lb,mb,rbfloat arrays

Left, mid and rigth-side of bins

f, ferrfloat arrays

Correlation function and its error

parMunch dictionary

Used to pass par.outfn to name the output file

fmtstring. Default=’%17.5f’

Numeric formatting string

altnamestring. Default=None

If supplied, use an alternative file name instead of par.outfn

Examples

import gundam as gun
c1 = gun.readcounts('redgals.cnt')

writeasc_cf(c1.rpl, c1.rpm, c1.rpr, c1.wrp, c1.wrperr, c1.par)
writeasc_rppicounts(lb, mb, rb, rppi, par, fmt='%17.5f', cntid=None, altname=None)[source]

Write an ASCII file for a rp-pi count array produced by the code. This is a 2D array of counts in the projected (rp) and radial (pi) directions.

The columns in the output will be [lb mb rb tot_counts rppi] where the first 3 are the left, mid and right-side of bins, tot_counts are the counts integrated for all radial bins, and rppi has one column for each radial bin

Parameters

lb,mb,rbfloat

Left, mid and rigth-side of bins

rppifloat array

2-dimensional count array. Usually this is one of the fields cnt.dd, cnt.rr, etc. of a projected correlation run

parMunch dictionary

Used to pass various data, including par.outfn to name the output file

fmtstring. Default=’%17.5f’

Numeric formatting string

cntidstring. Default=None

ID string for column headers. Usually can be ‘dd’, ‘rr’, ‘dr’, etc. Also appended as the extension of the ouput file (when altname=None)

altnamestring. Default=None

If supplied, use an alternative file name instead of par.outfn + .cntid

Examples

import gundam as gun
c1 = gun.readcounts('redgals.cnt')

c1.par.outfn
'/home/myuser/sdss/redgals'

# Write the DD counts in rp-pi dimensions
gun.writeasc_rppicounts(c1.rpl, c1.rpm, c1.rpr, c1.dd, c1.par, cntid='dd')

# Inspect the output file
with open('redgals.dd', 'r') as f:
    print(f.read(), end="")

#  lb        mb        rb        dd         dd_001    dd_002    dd_003    ...
#  0.10000   0.12417   0.14835   11509.00   2082.00   1500.00   1168.00   ...
#  0.14835   0.18421   0.22007   20273.00   3122.00   2378.00   1899.00   ...
#  0.22007   0.27327   0.32647   36169.00   4940.00   3845.00   3283.00   ...
#  0.32647   0.40539   0.48431   64866.00   8453.00   6302.00   5236.00   ...
#  ...
writeasc_counts(lb, mb, rb, c, par, fmt='%17.5f', cntid=None, altname=None)[source]

Write an ASCII file for a (1-dimensional) counts array produced by the code.

The columns in the output will be [lb mb rb c] where the first 3 are the left, mid and right-side of bins and c are the counts

Parameters

lb,mb,rbfloat

Left, mid and rigth-side of bins

cfloat array

Counts array. Usually this is one of the fields cnt.dd, cnt.dr, cnt.rr, etc. of a correlation run

parMunch dictionary

Used to pass par.outfn to name the output file

fmtstring. Default=’%17.5f’

Numeric formatting string

cntidstring. Default=None

ID string for column header. Usually can be ‘dd’, ‘rr’, ‘dr’, etc. Also appended as the extension of the ouput file (when altname=None)

altnamestring. Default=None

If supplied, use an alternative file name instead of par.outfn + .cntid

Examples

import gundam as gun
c1 = gun.readcounts('bluegals.cnt')

# Write the DD counts in angular dimensions. Use an alternative file name
gun.writeasc_counts(c1.thl, c1.thm, c1.thr, c1.dd, c1.par, cntid='dd', altname='akounts')

# Inspect the output file
with open('akounts', 'r') as f:
    print(f.read(), end="")

# lb        mb        rb           dd
# 0.01000   0.01206   0.01413     3178.00
# 0.01413   0.01704   0.01995     6198.00
# 0.01995   0.02407   0.02818    12765.00
# 0.02818   0.03400   0.03981    24888.00
# 0.03981   0.04802   0.05623    49863.00
# 0.05623   0.06783   0.07943    98883.00
...

Fortran Wrapper Routines

These are the wrappers that call Fortran pair counting routines. They are intended to: (1) choose the fastest counting routine depending whether weights, bootstrap errors, etc. are requested or not; and (2): concentrate all mayor Fortran calling in one place

Unless you are building your own custom script, heavily modifying, or extending the core algorithms, you normally should not need to use these functions

pairs_auto(par, wunit, logff, tab, x, y, z, sk, ll, dc=None)[source]

Wrapper for calling Fortran counting routines (for auto-pairs)

This function isolates the call of external Fortran routines, choosing the correct one based on geometry, and choosing the fastest depending whether, weights, bootstrap errors, etc. are requested or not.

Parameters

parMunch dictionary

Input parameters

wunitbool

  • True : all weights are equal to 1

  • False : at least one weight is not 1

logffstring

File for logging the output of Fortran routines

tabastropy table

Table with data particles

x,y,zarrays

Rectangular coordinates of data particles

sk,llarrays

Skip table (SK) and linked list table (see gundam.skll2d() or gundam.skll3d())

dcarray [optional]

Array of comoving distances. Not needed for angular counts

Returns

ttlist of ndarray

Ouput counts as returned by Fortran counting routines

pairs_cross(par, wunit, logff, tab, x, y, z, tab1, x1, y1, z1, sk1, ll1, dc=None, dc1=None)[source]

Wrapper for calling Fortran counting routines (for cross-pairs among two tables called D(data) and Random(R))

This function isolates the call of external Fortran routines, choosing the correct one based on geometry, and choosing the fastest depending whether, weights, bootstrap errors, etc. are requested or not.

Parameters

parMunch dictionary

Input parameters

wunitbool

  • True : all weights are equal to 1

  • False : at least one weight is not 1

logffstring

File for logging the output of Fortran routines

tab,tab1astropy tables

Table with data particles (D) and random particles (R), respectively

x,y,z,x1,y1,z1arrays

Rectangular coordinates of particles in D table and R rable, respectively

tab1astropy table

Table with data particles (D)

sk1,ll1arrays

Skip table (SK) and linked list table (see gundam.skll2d() or gundam.skll3d())

dc,dc1arrays [optional]

Array of comoving distances of particles in D and R tables. Not needed for angular counts

Returns

ttlist of ndarray

Ouput counts as returned by Fortran counting routines