==========================================================
Monte Python 2cosmos, a modification of Monte Python,
a Monte Carlo Markov Chain code (with CLASS!)
===========================================================
:Main developers: Fabian Koehlinger <fabian.koehlinger@ipmu.jp> (2cosmos),
Thejs Brinckmann <brinckmann@physik.rwth-aachen.de>,
Nils Schoeneberg <schoeneberg@physik.rwth-aachen.de> (Python3)
:Author: Benjamin Audren <benjamin.audren@epfl.ch>
:License: MIT
This modification of Monte Python allows the user to write likelihood modules
that can request two independent instances of CLASS and separate dictionaries
and structures for all cosmological and nuisance parameters. The intention is
to be able to evaluate two independent cosmological calculations and their
respective parameters within the same likelihood. The use case for this is the
evaluation of a likelihood using correlated datasets (e.g. mutually exclusive
subsets of the same dataset for which one wants to take into account all
correlations between the subsets). An example for such a use case is the
analysis of `Koehlinger et al. 2019 MNRAS, 484, 3126) <https://ui.adsabs.harvard.edu/abs/2019MNRAS.484.3126K>`
for which this modification of Monte Pythoin was originally set up. The '2cosmos'
likelihood of this analysis ('montepython/likelihoods/kids450_cf_2cosmos_likelihood_public')
is provided with this public release and shall serve as an example for writing
your own '2cosmos' likelihoods (refer also to the README of that likelihood).
The code is under the MIT license. As an additional clause, when using the code
in a scientific publication you are also required to cite ``A Bayesian quantification
of consistency in correlated data sets`` in addition to the Monte Python v3.0 release
paper ``MontePython 3: boosted MCMC sampler and other features`` and the original release
paper ``Conservative Constraints on Early Cosmology`` (see the tail of this document
for the bibtex entries).
IMPORTANT
---------
We have only tested Monte Python's default Metropolis-
Hastings sampler, the MultiNest and PolyChord samplers. If you want to use
the CosmoHammer sampler, please be aware of potentially occuring bugs (all
modifications related to its 2cosmos sampling capabilities are included in the
module).
For setting up this modified version of Monte Python and general instructions,
please follow the descriptions from the original README distributed with Monte
Python as copied below:
Copy of original Monte Python v3.0 README
-----------------------------------------
Details and Examples
--------------------
If you are searching for further details or specific examples of a work session,
please refer to the online documentation. See also the `Monte Python 3 paper
<https://arxiv.org/abs/1804.07261>`_ for details on the code, including a
summary of features as of v3.0.
Note the `Monte Python 3 paper <https://arxiv.org/abs/1804.07261>`_ contains an
overview of all likelihoods currently implemented in the code, with some details
on those likelihoods, such as datasets, last updated, type and relevant papers
to cite when using the likelihood. In the future, the overview of likelihoods
will be maintained on the official `Monte Python website
<https://brinckmann.github.io/montepython_public/>`_.
You can find installation details below and on the archived `Monte Python 2 wiki
<https://github.com/baudren/montepython_public/wiki>`_. The `Monte Python 3 forum
<https://github.com/brinckmann/montepython_public/issues>`_ contains a
collection of already answered questions, and can be used to discuss the code.
Also refer to the archived `Monte Python 2 forum
<https://github.com/baudren/montepython_public/issues>`_ for additional
previously answered questions, but please post all new issues on the
`Monte Python 3 forum <https://github.com/brinckmann/montepython_public/issues>`_.
The official `Monte Python website
<https://brinckmann.github.io/montepython_public/>`_, the
`course page of Julien Lesgourgues <https://lesgourg.github.io/courses.html>`_,
and the `hi_class website <http://miguelzuma.github.io/hi_class_public>`_ contain *Monte Python*
(and *Class*) lectures, examples and exercises.
Want to contribute?
------------------
*Monte Python* is developed and maintained by volunteer workers and we are always
happy for new people to contribute. Do not hesitate to contact us if you believe
you have something to add, this can be e.g. new likelihoods, new samplers,
improvements to the plotting, bug fixes, or ideas for how to improve the code.
Additionally, everyone is encouraged to assist in resolving issues on the forum,
so do not hesitate to reply if you think you can help.
In particular, if you would like to have your likelihood added to the public
github repository, please make sure it is well documented and add all relevant
information to the .data file, e.g. authors and references.
Prerequisites
-------------
* You need the python program **version 2.7** or above, but less than 3.0.
Note that lower versions of python will work, down to 2.6 (tested), if you
add manually two extra packages
(`ordereddict <http://code.activestate.com/recipes/576693/>`_ and
`argparse <https://pypi.python.org/pypi/argparse/1.2.1>`_).
* Your python of choice must have `numpy` (version >= 1.4.1) and `cython`. The
later is used to wrap CLASS in python.
* *[optional]* If you want to use fully the plotting capabilities of Monte Python,
you also need the `scipy`, with interpolate, and `matplotlib` modules.
* *[optional]* You can now use Multi Nest and the CosmoHammer with Monte
Python, though you need to install them. Please refer to the documentation.
The MontePython part
--------------------
Move the `.tar.bz2` file to the place of your convenience, untar its content
.. code::
$ bunzip2 montepython-vx.y.tar.bz2
$ tar -xvf montepython-vx.y.tar
This will create a directory named `montepython` into your current directory.
You can add the following line to your `.bashrc` file:
.. code::
export PATH=/path/to/MontePython/montepython/:$PATH
to be able to call the program from anywhere.
You will need to adapt only two files to your local configuration. The first
is the main file of the code `montepython/MontePython.py`, and it will be the only
time you will have to edit it, and it is simply to accommodate different
possible configurations of your computer.
Its first line reads
.. code::
#!/usr/bin/python
This should be changed to wherever is your preferred python distribution
installed. For standard distribution, this should already be working. Now,
you should be able to execute directly the file, i.e. instead of calling:
The second file to modify is located in the root directory of Monte Python :
`default.conf`. This file will be read (and stored) whenever you execute the
program, and will search for your cosmological code path, your data path, and
your wmap wrapper path. You can alternatively create a second one, `my.conf`,
containing your setup, and then run the code providing this file (with the flag
`--conf`)
The Class part
--------------
Go to your class directory, and do **make clean**, then **make**. This builds the
`libclass.a`, needed for the next step. From there,
.. code::
$ cd python/
$ python setup.py build
$ python setup.py install --user
This will compile the file `classy.pyx`, which is the python wrapper for CLASS,
into a library, `classy.so`, located in the `build/` subdirectory. This is the
library called in Monte Python afterwards.
If this step fails, check that you have `cython` installed, `numpy` (a numerical
package for python), python (well... did I say this code was in python ?) with
a version > 2.6. If this step fails again, kindly ask your system admin, (s)he
is there for this, after all. Note that the installation (last command) is
not strictly speaking mandatory.
Remember that if you modify `CLASS` to implement some new physics, you will need to
perform this part again for the new `CLASS`.
The Planck likelihood part
---------------------------
The release of the Planck data comes with a likelihood program, called
Clik, that one can recover from the `ESA website
<https://pla.esac.esa.int/pla/#cosmology>`_,
along with the data. Download all `tgz` files, extract them to the
place of your convenience.
The Planck Likelihood Code (**plc**) is based on a library called
`clik`. It will be extracted, alongside several `.clik` folders that
contain the likelihoods. The installation of the code is described in
the archive, and it uses an auto installer device, called `waf`.
.. warning::
Note that you **are strongly advised** to configure `clik` with the
Intel mkl library, and not with lapack. There is a massive gain in
execution time: without it, the code is dominated by the execution
of the low-l polarisation data from WMAP.
Go to your plc folder, and execute the following line, taking into
account the mkl installation
.. code::
$ ./waf configure --install_all_deps --mkl=...
In your |MP| configuration file, to use this
code, you should add the following line
.. code:: python
path['clik'] = 'path/to/your/plc/folder/'
The four likelihoods defined in |MP| for Planck are `Planck_highl`,
`Planck_lowl`, `Planck_lensing`, `lowlike` (the polarization data from
WMAP). In each of the respective data files for these likelihood,
please make sure that the line, for instance,
.. code:: python
Planck_highl.path_clik = data.path['clik']+'../something.clik'
points to the correct clik file. Do not forget to source your Planck
likelihood every time you want to use it:
.. code::
$ source Your/Plc/bin/clik_profile.sh
You can put this line in your .bashrc file, and you should put it in your
scripts for cluster computing.
Enjoying the difference
-----------------------
Now the code is installed. Go anywhere, and just call
.. code::
$ python montepython/MontePython.py --help
$ python montepython/MontePython.py run --help
$ python montepython/MontePython.py info --help
To see a list of all commands. For the `run` subcommand, there are two
essential ones, without which the program will not start. At minimum, you
should precise an output folder (`-o`) and a parameter file (`-p`). An example
of parameter file is found in the main directory of MontePython (`test.param`,
for instance).
A typical call would then be:
.. code::
$ python montepython/MontePython.py run -o test -p example.param
If non existent, the `test/` folder will be created, and a run with the number
of steps described in `example.param` will be started. To run a chain with more
steps, one can type:
.. code::
$ python montepython/MontePython.py run -o test -p example.param -N 100
If you want to analyse the run, then just type
.. code::
$ python montepython/MontePython.py info test/
Note that you probably want more than a hundred points before analyzing a
folder.
Bibtex entry
------------
When using *Monte Python* in a publication, please acknowledge the code by citing
the following papers. If you used *Class*, *MultiNest*, *PolyChord* or *Cosmo Hammer*,
you should also cite the original works.
Please also cite the relevant papers for each likelihood used: as of v3.0 we have a
list of references for all likelihoods in the first of the papers below. In the
future the list will be maintained on the official `Monte Python website
<https://brinckmann.github.io/montepython_public/>`_. Otherwise, this information can
often be found in the .data file of the likelihood folder.
In order to encourage people to both develop and share likelihoods with the community,
to the benefit of all users, we optionally encourage users to cite the paper in which
the *Monte Python* likelihood was first used, in addition to the papers in which data
and/or likelihoods were published.
.. code::
@article{Koehlinger:2019,
author = {{K{\"o}hlinger}, Fabian and {Joachimi}, Benjamin and {Asgari}, Marika and {Viola}, Massimo and {Joudaki}, Shahab and {Tr{\"o}ster}, Tilman},
title = "{A Bayesian quantification of consistency in correlated data sets}",
journal = {\mnras},
keywords = {gravitational lensing: weak, methods: data analysis, statistical, cosmology: cosmological parameters, observations, large-scale structure of Universe, Astrophysics - Cosmology and Nongalactic Astrophysics},
year = "2019",
month = "Apr",
volume = {484},
number = {3},
pages = {3126-3153},
doi = {10.1093/mnras/stz132},
archivePrefix = {arXiv},
eprint = {1809.01406},
primaryClass = {astro-ph.CO},
adsurl = {https://ui.adsabs.harvard.edu/abs/2019MNRAS.484.3126K},
adsnote = {Provided by the SAO/NASA Astrophysics Data System}
}
@article{Brinckmann:2018cvx,
author = "Brinckmann, Thejs and Lesgourgues, Julien",
title = "{MontePython 3: boosted MCMC sampler and other features}",
year = "2018",
eprint = "1804.07261",
archivePrefix = "arXiv",
primaryClass = "astro-ph.CO",
SLACcitation = "%%CITATION = ARXIV:1804.07261;%%"
}
@article{Audren:2012wb,
author = "Audren, Benjamin and Lesgourgues, Julien and Benabed,
Karim and Prunet, Simon",
title = "{Conservative Constraints on Early Cosmology: an
illustration of the Monte Python cosmological parameter
inference code}",
journal = "JCAP",
volume = "1302",
pages = "001",
doi = "10.1088/1475-7516/2013/02/001",
year = "2013",
eprint = "1210.7183",
archivePrefix = "arXiv",
primaryClass = "astro-ph.CO",
reportNumber = "CERN-PH-TH-2012-290, LAPTH-048-12",
SLACcitation = "%%CITATION = ARXIV:1210.7183;%%",
}
fkoehlin/montepython_2cosmos_public
Public repository for the Monte Python Code modified to work with two independent cosmological parameter sets and independent cosmological calculations per likelihood.
PythonNOASSERTION