PyMC: Bayesian Stochastic Modelling in Python

By Anand Patil, David Huard, Christopher Fonnesbeck. Journal of Statistical Software, 2010

705 citations recorded by the source corpus.

This user guide describes a Python package, PyMC, that allows users to efficiently code a probabilistic model and draw samples from its posterior distribution using Markov chain Monte Carlo techniques.

Subfields: Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/9762b7357c0deb9208ace1f6951ab7fa2098535c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

Citations in the corpus

PaCAL: A Python Package for Arithmetic Computations with Random Variables (2014, background): “The MC approach currently oﬀers the greatest ﬂexibility allowing for an arbitrary number of dependent random variables, Bayesian inference, etc. Examples of such systems are BUGS (Lunn, Thomas, Best, and Spiegelhalter 2000; Spiegelhalter, Thomas, Best, and Lunn 2003) and PyMC (Patil, Huard, and Fonn”
The Challenge of Machine Learning in Space Weather: Nowcasting and Forecasting (2019, background): “Some examples of community-developed python libraries that are now widely used are theano (Bergstra et al., 2010), scikit-learn (Pedregosa et al., 2011), astroML (VanderPlas et al., 2012), emcee (VanderPlas et al., 2012), and PyMC (Patil et al., 2010), among many others.”
Predictive Entropy Search for Bayesian Optimization with Unknown Constraints (2015, methodology): “, 2006) to compute the effective sample size and the Geweke convergence diagnostic, and the PyMC python package (Patil et al., 2010) to compute the Gelman-Rubin diagnostic over two independent traces.”
Global numbers of infection and disease burden of soil transmitted helminth infections in 2010 (2014, neutral): “Carlo (MCMC) algorithm [29].”
The global distribution and burden of dengue (2013, neutral): “Asia bore 70% (67 [47-94] million infections) of this burden, and is characterised by large swathes of densely populated regions coinciding with very high suitability for disease transmission.”
HDDM: Hierarchical Bayesian estimation of the Drift-Diffusion Model in Python (2013, methodology): “HDDM is an open-source software package written in Python which allows (1) the ﬂexible construction of hierarchical Bayesian drift diffusion models and (2) the estimation of its posterior parameter distributions via PyMC (Patil et al., 2010).”
Probing the Standard Model with Higgs signal rates from the Tevatron, the LHC and a future ILC (2014, methodology): “For an efficient sampling of the parameter space the scans are performed with an adaptive Metropolis (AM) algorithm [90] with flat prior probability distributions using the Markov-Chain Monte Carlo (MCMC) python package PyMC [91].”
Discontinuous Hamiltonian Monte Carlo for Probabilistic Programs (2018, background): “We compare against NUTS (Hoffman & Gelman, 2014) combined with Metropolis-within-Gibbs (NUTS+MwG) in PyMC3 (Salvatier et al., 2016) and Random-walk Lightweight Metropolis-Hastings (RMH) in Anglican (Wood et al., 2014).”
State-space models for bio-loggers: A methodological road map (2013, background): “Software such as WinBUGs (Lunn et al., 2000), OpenBUGs (Lunn et al., 2009), JAGs6 and PyMC (Patil et al., 2010) have made MCMC highly accessible.”
Predicting human behavior in unrepeated, simultaneous-move games (2013, methodology): “• Stahl and Wilson (1995) evaluated prediction performance on 3 games using parameters fit from the other games; • Morgan and Sefton (2002) and Hahn et al. (2010) evaluated prediction performance using held-out test data; • Camerer et al.”