dive documentation

dive is a Python package for Bayesian analysis of dipolar EPR (electron paramagnetic resonance) spectroscopy data through Markov chain Monte Carlo (MCMC) sampling with the Python package PyMC.

About

dive is a Python package for analyzing DEER spectroscopy with Bayesian inference. When given a model (Tikhonov regularization or Gaussian mixture), prior distributions for model parameters (which are already included in dive), and experimental data, Markov chain Monte Carlo (MCMC) sampling is run with PyMC to yield refined posterior distributions for each parameter. These posterior distributions provide full quantification of the uncertainty of all parameters, and can be analyzed for their means, confidence intervals, and so on.

Usage

See Usage for information on how to use dive.

Contents

• Usage
  • Installation
  • Reading Data
  • Modeling
  • Sampling
  • Saving
  • Validating
  • Accessing Results
  • Plotting
• Reading & Creating Data
  • Reading .DSC Data
    • deerload()
    • read_description_file()
  • Creating Synthetic Data
    • generate_multi_gauss()
    • generate_single_gauss()
  • Creating P(r) and B(t)
    • bg_exp()
    • bg_hom3d()
    • dd_gauss()
    • gauss()
• Modeling
  • Creating a Model
    • model()
  • PyMC Models
    • regularizationmodel()
    • multigaussmodel()
• Sampling
  • Sampling a Model
    • sample()
  • Backend Samplers
    • DeltaSampler
      • DeltaSampler.step()
    • PSampler
      • PSampler.step()
    • TauSampler
      • TauSampler.step()
• Plotting
  • draw_posterior_samples()
  • pairplot_chain()
  • pairplot_condition()
  • pairplot_divergence()
  • plot_P()
  • plot_V()
  • plot_correlations()
  • plot_hist()
  • plot_marginals()
  • print_summary()
  • summary()
• Saving
  • load_trace()
  • save_trace()
• Utility Functions
  • FWHM2sigma()
  • addnoise()
  • dipolarkernel()
  • fnnls()
  • get_rhats()
  • prune_chains()
  • sigma2FWHM()