Quickstart

In this section we show a basic four steps to use SimpleMC:

  1. Set an ini file.

  2. Read the configuration in a Python script.

  3. Then run in terminal.

  4. Finally analyze outputs.

ini file

The ini file has all the necessary options to SimpleMC, the mandatory options are in the [custom] section, the rest have default values corresponding to specific analyzers and you can modify them accordingly with your needs (see Customize inifile or baseConfig.ini file for more information).

The [custom] section has the following structure:

[custom]

chainsdir = chains

model = LCDM

datasets = BBAO+HD+SN

analyzer = mcmc

you must choose an existing directory to save the outputs (chains, summary, and .paramnames). The options for model, datasets and analyzer are as follows:

Note

  • model: visit Models section to see the options.

  • analyzer options:

    • mcmc, nested, emcee, MaxLikeAnalyzer, genetic, ga_deap

  • data options (you can combine any of them): visit Data section to see the options.

Python script

We can use test.py with the path of the ini file:

from simplemc.DriverMC import DriverMC

fileConfig = "path/baseConfig.ini"
D = DriverMC(fileConfig)

run in terminal

For last, run in the terminal:

$ python test.py

parallelization

To run multiple MCMC (Metropolis-Hastings) chains in parallel:

mpirun -np 4 python test.py

where 4 is the number of chains and the number of processors.

The nested sampling algorithms use the multiprocessing library, therefore it is only necessary to set in the [nested] section of the ini file the number of cores with the nproc key:

[custom]
...
...
analyzer = nested
...

[nested]
...
; using 3 cores:
nproc = 3
...

analyze outputs

You can see the outputs in the chains directory and then make plots. See the plots section for details. The name of the outputs begins with the name of the model, prefact (pre / phy), datasets and analyzer, for the example of the above ini file we have the output name: LCDM_phy_BBAO+HD+SN_mcmc.

In addition to the chain file, it is a summary where you can notice the parameter estimation, the execution time and in the case of nested sampling, the Bayesian evidence, useful for the comparison of models.