Structure of the repository

The darkmix repository contains the code of the project, the sample data used in the Publication, as well as the output files obtained in that analysis.

Code

The folder Code contains the following files:

  • darkmix.R: contains the R functions used in the library. It must be loaded every time the library is used.

  • kernel_lambda.cpp: C++ code for the kernel density calculations. Not meant to be used directly by the user.

  • kernel_absolute_residuals.cpp: C++ code for the residuals kernel density estimation. Not meant to be used directly by the user.

  • kernel_fourier_residuals.cpp: fast C++ code for the residuals kernel density estimation. Not meant to be used directly by the user. Requires the installation of the FFTW library (see Installation).

  • darkmix_steps.Rmd: notbook containing the execution steps as used in the publication. Uses the data stored in folder Data. You can see this example of execution in Example of execution.

Data

The folder Data contains the following files:

  • datacat.txt: data set containing the coordinates of 2081 dark matter particles from the Bolshoi simulation . Used in the darkmix_steps.Rmd example code.

  • halocat.txt: catalog of halos found by the BDM algorithm in our region of interest.

  • bdm_halos.txt: selection of 10 halos from the halocat.txt catalog coincident with the ones found in the datacat.txt sample following darkmix_steps.Rmd.

Output

The folder Output contains the following files:

  • parameters.txt: best fit paramters of our 11-component mixture model. The last row corresponds to the background component and column N is the number of estimated particles per component.

  • membership.txt: probability for each of the 2081 particles in datacat.txt of belonging to each of the 11 components. Column class is the component to which the particle has been assigned in our example.