README File for SDSS-DR9 CMASS Galaxy Clustering patch for CosmoMC May2010 Written by Antonio Cuesta, Kushal Mehta, Beth Reid and Lado Samushia, June 2012 ========================================================= This file implements into CosmoMC the clustering measurements from the correlation function of the CMASS Data Release 9 galaxy sample of SDSS-III/BOSS, with a median redshift of z=0.57. The two main results encoded in this patch are the measurement of the position of the baryon acoustic oscillation feature (hereafter BAO), and the shape of the monopole and quadrupole moments (hereafter anisotropic clustering), both for the CMASS-DR9 correlation function. In addition, it includes the recent re-analysis of the position of the BAO in the LRG sample from Data Release 7, with a median redshift of z=0.35. The BAO measurement makes use of the usual angle-averaged distance estimator D_V(z)=[(1+z)^2*DA(z)^2*c*z/H(z)]^(1/3), where DA(z) is the angular diameter distance and H(z) is the Hubble parameter. The anisotropic measurement generalizes the usual angle-averaged distance estimator D_V(z)=[(1+z)^2*DA(z)^2*c*z/H(z)]^(1/3), with two additional estimators that measure the degree of line-of-sight anisotropies in galaxy clustering, i.e. we simultaneously measure the Alcock-Paczynski parameter F(z)=(1+z)*DA(z)*H(z)/c and the redshift-space distortion parameter measuring the growth rate of structures: d(sigma8(z))/d(-ln(1+z))=f(z)*sigma8(z) where f(z) is the logarithmic derivative of the linear growth function and sigma8(z) is the rms density fluctuation within a top-hat sphere of 8Mpc/h radius, assuming linear growth. In the case of anisotropic clustering, the three measurements above are combined into a single multivariate gaussian likelihood, which if used in combination with Cosmic Microwave Background and Type Ia Supernovae data, it improves the constraints on the LCDM cosmological model and extensions of it, in particular on cosmologies with free curvature. The CMASS DR9 BAO measurement is also approximated by a Gaussian likelihood function. This is not the case for the LRG DR7 BAO measurement since the likelihood is somewhat skewed, so the likelihood is computed using the file "data/sdss_baoDR7.txt" included in this patch. For more details regarding the BAO measurements please refer to: http://arxiv.org/abs/1203.6594 (CMASS DR9, Anderson et al. 2012) http://arxiv.org/abs/1202.0090 (LRG DR7, Padmanabhan et al. 2012) For more details regarding the anisotropic clustering measurement please refer to: http://arxiv.org/abs/1203.6641 (CMASS DR9, Reid et al. 2012) http://arxiv.org/abs/1206.5309 (CMASS DR9, Samushia et al. 2012) HOW TO INSTALL IT: ================== The attached patch file ("CMASSDR9_Reid_cosmoMC_multiBAO.patch") modifies the source files of CosmoMC (Lewis & Bridle 2002) to include the measurements described in the previous section. Note that this patch was generated to work with the May 2010 version of CosmoMC (the latest version as of this README file was written). We do not guarantee that this patch will work with future versions of CosmoMC. Note however, that it is possible to download the May 2010 version from http://cosmologist.info/cosmomc/ We also cannot guarantee that the patch will work on previously patched versions of CosmoMC (but see below). 0) Replace the file "camb/equations.f90" with the file "equations_cross.f90" in the PPF module by Wenjuan Fang. To download it, go to http://camb.info/ppf/ (version 2008). This is a requirement since our patch requires that the equation of state of dark energy is parameterized as w(a)=w_0+(1-a)*w_a, and the PPF module provides support for this. 1) Copy the attached patch file ("CMASSDR9_Reid_cosmoMC_multiBAO.patch") to the root directory of CosmoMC. This is the directory that contains the executable file, and is one level above the 'source' and 'data' directories. 2) Go to that directory and run the following: patch -p1 -i CMASSDR9_Reid_cosmoMC_multiBAO.patch 3) Go to the 'source' directory and overwrite your Makefile with the provided one (if you had previously changed your Makefile, you will need to include those changes again, remember to make backups): cp Makefile_sdssdr9multi Makefile 4) Run 'make' to compile the new code and create the executable. (Note: you need to have the GNU Scientific Library installed, otherwise the code will not compile). If this is the first time that you compile CosmoMC, note that you will have to link the WMAP7 data files to your 'data' directory and you will also have to modify the Makefile in 'source' to set the location of the WMAP7 likelihood code and the cfitsio library. Please go to http://cosmologist.info/cosmomc/readme.html for more details. If you want to install the CosmoMC module from the 3-year Supernovae Legacy Survey (SNLS3, Conley et al. 2011), which is the supernova dataset used in Anderson et al. 2012, download the patch and follow the instructions available at: http://casa.colorado.edu/~aaconley/cosmomc_snls/ Please note that if you plan to use the SNLS3 patch in combination with this one, you will need to apply CMASSDR9_Reid_cosmoMC_multiBAO.patch *first* and then when you apply the SNLS3 patch you need to increase the fuzz factor, otherwise some files will fail to be patched. To do this, include the -Fn option, with n >= 3: patch -F3 -p1 -i add_snls.patch Then, overwrite your Makefile with the file "Makefile_sdssdr9multi_snls" included in the patch (if you had modified your Makefile before, you need to include those changes to Makefile_sdssdr9multi_snls before overwriting your Makefile, always make backups!). Also remember to modify your "params_CMB.paramnames" and the "params_CMB.covmat" files accordingly. For the latter, you can use the file "params_CMB_alphabeta.covmat" included in the patch. HOW TO USE IT: ============== In order to include the anisotropic clustering measurements in your cosmological analysis, make sure that in your parameter file (typically "params.ini") you have set these two flags to TRUE: get_sigma8 = T use_BAO_DR9_ANISO = T Additionally, if you wish to include the isotropic BAO measurements, make sure that in your parameter file (typically "params.ini") you have set one or both of these two flags to TRUE: use_BAO_DR7 = T use_BAO_DR9 = T Optionally, as used in Anderson et al. 2012 it is possible to include the low redshift BAO measurements from the 6dFGS survey (Beutler et al. 2011). To use these measurements, set this flag to TRUE: use_BAO_6DF = T To use the latest Hubble constant measurements, we have replaced the "HST.f90" module with a similar file in which we implemented the results from Riess et al. (2011, ApJ, 730, 119). As in the original version of CosmoMC, to use these constraints set this flag to TRUE: use_HST = T Then run the code as usual. Note that in order to avoid using the same dataset twice, the DR7 BAO measurements cannot be used simultaneously with the DR7LRG data. Likewise, using the DR9 BAO and the DR9 anisotropic clustering data will result in error. It is possible, however, to use the DR7 BAO and any of the DR9 BAO datasets at the same time, since the overlap between both samples is negligible, so they can be treated as independent samples. README file. Last modified: September 15, 2012