SimpleMC

[Site under construction] [Last update 30/08/2023]

This is the SimpleMC documentation, a Python package for cosmological parameter estimation and model comparison using Bayesian inference, optimization and machine learning algorithms. SimpleMC also contains a collection of tools for cosmological data analysis, such as a cosmological calculator, plotting utilities and other statistical issues. The cosmological models included in SimpleMC are Dark Energy models where only expansion history matters. In the simplemc_workshop repository, there are some slides about SimpleMC (in spanish) and an example of a jupyter-notebook using SimpleMC in Google Colab.

There are 3 options to install, or use, SimpleMC:

1. You can use SimpleMC in Google Colab, please see the Section Using Google Colab in Requirements.

2. On the other hand, you can visit the GitHub code repository, download the source code here or clone it as follows:

   git clone https://github.com/ja-vazquez/SimpleMC.git
   

Then, you can install it:

cd SimpleMC

pip3 install -e .

3. Also, you can install SimpleMC without clonning:

   pip3 install -e git+https://github.com/ja-vazquez/SimpleMC#egg=simplemc
   

In a few days, the pip install option will be available pip3 install simplemc.

Please read the introduction section where you can see the requirements and a simple quick start. In the tutorials section you can find several examples of SimpleMC functions.

SimpleMC employs amazing external codes and algorithms, please go to the citation section for details if you feel it is pertinent to cite them.

Extended index

• First steps
  • Requirements
    • Installing dependencies in python
    • Using Google Colab
  • Quickstart
    • ini file
    • Python script
    • run in terminal
    • parallelization
    • analyze outputs
  • Customize ini file
    • baseConfig.ini
• Structure
  • Data
    • Datasets combinations
  • Models
  • Analyzers
    • Sampler comparison
  • General structure
• Tutorials
  • Cosmological calculator
    • Notebook example
  • Set a prior probability density function
  • How to create a new model
    • Simple model independent of any cosmology
    • Simple cosmological model based on LCDM
  • How to use an external dataset?
    • From ini file with the existing cosmological functions
    • Without cosmological functions
  • Bayesian inference
    • With Metropolis-Hastings algorithm
    • With EMCEE algorithm
    • With nested sampling
    • Notebook example
  • Maximum Likelihood Estimation (MLE) with optimizer
    • Notebook example
  • MLE with genetic algorithms
    • Notebook example
  • Model comparison with Bayesian evidence
    • Notebook example
  • Plot the outputs
    • With getdist:
    • With Corner:
    • With Simple_Plot class
    • With fgivenx
  • Toy models
  • Speed-up Bayesian inference with neural networks
    • Neuralike
• References
  • Cite SimpleMC
  • Cite external codes
  • Some research using SimpleMC
• API
  • Top level interface
  • Likelihood
  • Cosmo
  • Models
  • Analyzers
  • Plots

Changelog

• 0.9.8.9 (28/06/2023) New cosmic chronometers added (HD23).

• 0.9.8.8 (28/06/2023) New BAO data added (DR16BAO).

• 0.9.8.7 (27/06/2023) Pantheon Plus added (PantheonPlus).

• 0.9.8.6 (24/05/2022) Fix MCMCAnalyzer and update requirements.

• 0.9.8.6 (27/04/2022) Fix mpi4py import in MCMCAnalyzer and allow to work it with only 1 processor.

• 0.9.8.5 (23/02/2022) Improving postprocessing.

• 0.9.8.4 (12/01/2022) Unifying binned and full Pantheon likelihoods. Added covariance matrix in postprocessing. Added AIC information criterion. Saving output of individuals from GADeap.

• 0.9.8.3 (17/12/2021) Fix binned Pantheon data and likelihood. Save maxloglikes in postprocessing. Reordering optimizers.

• 0.9.8.2 (04/11/2021) Update paths of datasets, setup and requirements.

• 0.9.8.1 (06/09/2021) Setup file and pypy install ready. Data in likelihood modules use relative paths.

• 0.9.8 (06/09/2021) Improvements in postprocessing. Return dictionaries in DriverMC.executer. Correction in the reading of the columns by mcevidence.

• 0.9.7 (02/09/2021) Add emcee and mcevidence source codes (modified). Fix output texfile of ga_deap. Delete old genetic algorithm.

• 0.9.6.8 (12/08/2021) Summary output to all analyzers. Remove nestle engine.

• 0.9.6.5 (09/08/2021) Add ga_deap options in ini file.

• 0.9.6 (7/06/2021) Unify simple and generic models, test generic likelihood.

• 0.9.5.8 (26/04/2021) Quintom with coupling and curvature. Beta version of neuralike. Warnings for non-essential libraries.

• 0.9.5.4 (21/03/2021) Fixed parameter estimation problem in Summary of nested chains.

• 0.9.5.2 (07/02/2021) GA from deap working with fisher matrix for errors.

• 0.9.5 (19/09/2020) Test and fix bambi with modified dynesty.

• 0.9.4 (25/06/2020) Working on neutrinos. New models: logt, IBEG, anisotropic, brans-dickie. Maxanalizer and fisher working.

• 0.9.3.5 (05/06/2020) SimpleMC Workshop via Zoom for ICF-UNAM

• 0.9.3 (02/06/2020) Fix typos and little errrors. DriverMC. Testing models and cosmology. Running and cleanning genetic functions.

• 0.9.2 (23/04/2020) adding modified dynesty. Postprocessing. Compressed data. GR for a single chain.

• 0.9.1 (06/04/2020) paralelise MCMC and added GR. Simple plots. Nested samplers. Updating hubble diagram to 31 points.

• 0.9.0 (28/11/2019) Gelman-Rubin diagnostics for MCMCAnalyzer. Ini files. Generic models and likelihoods.

• 0.8.9 (25/11/2019) Moving likelihoods into a likelihoods folder and models on to models folder.

• 0.8.7 (20/11/2019) Adding HD and full JLA. Ordering directories.

• 0.8.6 (20/08/2019) Bug in the MCMC sampler.

• 0.8.5 (29/04/2019) Updated Lya likelihood to DR14.

• 0.8.4 (23/03/2019) python 3, towards eBOSS.

• 0.8.3 (10/10/2017) Added binned cosmology.

• 0.8.2 (25/07/2017) DR12 and DR14 data.

• 0.8.1 (06/01/2016) Added sampling test for demonstration of MCMC.

• 0.8.0 (01/09/2014) Initial version.

TO DO

• Add nerualike to the main branch.

• Add Planck-18.