# cosmic_anisotropy

## Project Description
This repository contains the files related to the development of a codebase in Python which performs a simple Monte Carlo simulation to determine best fit parameters in an anisotropic model of the universe. It was a join effort between Wyatt Carbonell and Azadé Modaressi, replicated a Mathematica file developed by Michael Seifert, and uses simulated data generated by Chloe Stults and Tori Plaskon. Our goal is to document our work sufficiently so that a future researcher can improve upon our Monte Carlo. Many thanks to Yale University and the CAREERS Cyberteam program for their funding and computing resources.

## Contents of Folders
The following folders are available in this repository, and contain the listed information:
 * Data - 10 example outputs of running the simulation for a few hundred thousand Monte Carlo steps
 * Mathematica Files - Mathematica Notebooks including the original Monte Carlo which was replicated (MCMC Distribution ...) as well as a few quality control checks for comparison against our Python code (...TestABvacmetric0.nb and ...chi-squared testing code.nb) and a csv of results which can be loaded into the Python for comparison's sake
 * Simulation Files - Python code which implements the simulation, as well as the simulated data csv used. For more information about each file, see their respective docstrings and the Organization section.
 * Slurm Out Files - Generated by the Grace cluster at Yale during simulation execution

The remaining files include information about the miniconda environment used to execute the simulation (both human readable and explicit), and a bash script used to submit jobs to Grace.

 ## Dependencies
 - Numpy
 - Scipy
 - Matplotlib
 - Pandas
 - Argparse

 ## Organization
 This is organized from the top down, so files at the top depend on the files below. Levels are indicated by a number of *.

 *SingleMCMCRun.py will run a single Monte Carlo simulation for an inputed number of steps
 **MonteCarloSimulation.py will perform Monte Carlo steps one by one
 ***ChiSquared.py will calculate chi squared for a given set of parameters, as compared to a dataset object
 ****LoadingDataFromCSV.py will load in data from a csv into a dataset object
 ****DistanceModulus.py computes the predicted distance modulus for a given set of parameters
 *****HelperFunctions.py calculates the useful functions t, q, and psi
 ******ABvacmetric0.py calculates the functions A, B, C = A', and D = B', which are representative of cosmological expansion

 HelperGrapher.py was created solely to assist in debugging, and compares to the original Mathematica
 GeneratePoints.py can be used to generate a large number of start points at once, and remains untested as it was never used for anything.