Features - lopez86/PyWIMPs GitHub Wiki

Working

Astronomy and Particle Physics (astro and xsec modules)

Basic cross section model
Several form factors for both spin-independent and spin-dependent interactions
Standard halo model velocity distribution
Halo mean direction and velocity
Nucleus to nucleon normalization

Limit and interval setting (limits module)

Poisson frequentist upper limit (no background)
Feldman-Cousins confidence interval (known background rate)
CLs limits in a counting experiment.

Detector effects (det module)

Basic efficiency models: constant and logistic
Basic response models: ideal detector and Gaussian energy resolution
Experiment class to hold detector and physics models

Monte Carlo-based sampling (mc module)

Weighted samples using uniform velocity throws
Weighted samples drawing from a Maxwellian (Gaussian) velocity distribution
Unweighted samples using rejection sampling
Unweighted samples using a Markov chain with the Metropolis-Hastings algorithm

Note: All of these are designed to be used with the standard cross section and halo models here. They may not work with alternative models.

Examples

Some code to start training an MCMC model
Some code to compare the different sampling methods
An example of an annual modulation curve (threaded and unthreaded versions)

Future

Interactions

Inelastic dark matter scattering (small splitting between dark matter WIMPs and an excited state)
Coherent elastic neutrino scattering

Likelihood fitting of parameters
Bayesian limits
Maximum gap limits (Yellen 2002)
Model-independent annual modulation analysis
Spectrum and skymap plots
Cross section limit plots
Some ideas for daily modulation (directional) limits
Documentation - Explanation of physics, math, and stats used

Possible ambitious future project:

Toy simulation of a xenon TPC detector. Simulate a recoil and use geometry to assign hits to arrays of PMTs. Use machine learning techniques to calculate hit energies and positions (or at least a binary inside/outside fiducial volume)