#######################################################
                         PLC 1.0
#######################################################

This is the first public release of the Planck 
collaboration likelihood code and data.


Introduction
************

The package contains

* source code: 
    * plc-1.0.tar.bz2 (C, f90 and python likelihood library and tools)

* data
    * commander_v4.1_lm49.clik (low-ell TT likelihood)
    * lowlike_v222.clik (low-ell TE,EE,BB likelihood)
    * CAMspec_v6.2TN_2013_02_26.clik (high-ell TT likelihood)
    * lensing_likelihood_v4_ref.clik_lensing (lensing likelihood)

Building and installing the code
********************************

To build the code, you will need a lapack library and
a cfitsio library compiled with the repositionable option,
or as shared library. This is the case for the MKL or macos
lapack library, and also for cfitsio whose installer as an 
option for compiling as a shared library. 
Optionaly, extra python wrappers and tools can be installed 
if a suitable python 2.6+ distribution is available (including, 
numpy, cython and pyfits).

To build the code, one can either use the waf tool or modify
the included Makefile and use make. 

Detailed information on the building and installing process 
as well as the dependencies are given in the clik.pdf file
in the source package.

Once the building and installing step is fullfilled, you 
should have access to
libclik.so (the main library)
libclik_f90.so (the f90 wrapper of the main library)
which are the core tools.

Detailed information on the libraries, how to use them and 
the accompagning tools is provided in clik.pdf that can be 
found in the source package.

Using the likelihood
********************

Each likelihood file can be used with the library to compute
log likelihood values for an input power spectra and nuisance 
parameters. Each likelihood file requires a different set
of power spectra data (TT, TE, EE, BB, ell range) and a 
different set of nuisance parameters (foreground model, 
instrument model). 

The computation can either be done using the standalone executables, or 
calling the library. 
The standalone executable are 
clik_example_C and clik_example_f90 (optionally clik_example_py) 
and also serve as further examples of how to use the library. 

Some basic information on the likelihood data files
***************************************************

commander_v4.1_lm49.clik
------------------------

TT
lmax = 49
no extra parameters.

This is the low-ell TT Planck likelihood. It uses the 
Commander-Ruler Component separated results from the LFI and HFI maps.  

lowlike_v222.clik
-----------------

TE, EE, BB
lmax = 32
no extra parameters

This is the low-ell TE,EE, BB likelihood. It uses wmap9 data and code.
More precisely, the polarization maps are the wmap9 one, while the 
T map is from Planck (using the same Commander-Ruler data).
Beware than to compute this likelihood one has to provide not only
the TE, EE and BB model spectra but also the TT which is used in the
calculation of the covariance matrices.

CAMspec_v6.2TN_2013_02_26_dist.clik
------------------------------

TT
lmax 2500
14 extra parameters

This computes the Planck TT likelihood for the multipoles between 
50 and 2500. It uses data from the 100Ghz, 143Ghz and 217Ghz HFI 
channels. It needs 15 extra parameters describing the foregrounds and
instrumental model
  * A_ps_100  : the point source contribution at 100Ghz, \muK^2 at l=3000
  * A_ps_143  : the point source contribution at 143Ghz, \muK^2 at l=3000
  * A_ps_217  : the point source contribution at 217Ghz, \muK^2 at l=3000
  * A_cib_143 : the clustered cib contribution at 143Ghz, \muK^2 at l=3000
  * A_cib_217 : the clustered cib contribution at 217Ghz, \muK^2 at l=3000
  * A_sz      : the tSZ contriibution at 143Ghz
  * r_ps      : the correlation between A_ps_143 and A_ps_217
  * r_cib     : the correlation between A_cib_143 and A_cib_217
  * n_Dl_cib  : the slope of the clustered cib spectrum (Cl^cib ~ l^(2+n_Dl_cib)
  * cal_100   : the relative calibration of the 100Ghz channel to the 143Ghz
  * cal_217   : the relative calibration of the 217Ghz channel to the 143Ghz
  * xi_sz_cib : the correlation between SZ and cib
  * A_ksz     : The amplitude of the kSZ
  * Bm_1_1    : The first eigenmode of the beam error for the 100Ghz channel


lensing_likelihood_v4_ref.clik_lensing
--------------------------------------

phiphi
lmax 2048
no extra parameter

This computes the Planck lensing likelihood for the phiphi multipoles 
between 40 and 400. Beware that the computation requires both the phiphi
and TT power spectra.

Computing the full Planck likelihood
************************************

To compute the full planck likelihood, one need to add the result of 
the call of the likelihood code against each of the 4 likelihood files. 
This can be done either by calling each likelihood one after the other,
or by first grouping the CMB likelihoods (commander_v4.1_lm49.clik, 
lowlike_v222.clik and CAMspec_v6.2TN_2013_02_26.clik) in a composite 
likelihood file using the optional tool clik_join and then calling only 
this file and the the lensing one.
 

---
2013-03-08