marcizhu / PyStats

A Python 3 package for R-like statistical functions and distributions

PyStats

PyStats is a Python 3 library of statistical analysis and distribution functions with simple R-like syntax, scalar/list input/output with OpenMP parallelization.

Contents

• Distributions
• Installation
• Documentation
• Syntax and Examples
• Credits
• License

Distributions

Functions to compute the cdf, pdf, quantile, as well as random sampling methods, are available for the following distributions:

• Bernoulli
• Beta
• Binomial
• Cauchy
• Chi-squared
• Exponential
• F
• Gamma
• Inverse-Gamma
• Laplace
• Logistic
• Log-Normal
• Normal (Gaussian)
• Poisson
• Student's t
• Uniform
• Weibull

Installation

You can use pip to install this library:

pip3 install pystats

Alternatively, you can also clone this repository and install the plugin manually using pip:

$ git clone [e-mail]:marcizhu/PyStats.git
$ pip3 install ./PyStats

After that, you can just import pystats (or do from pystats import * if you don't want to type pystats. before all functions) and you're ready to go.

Documentation

Full documentation is available online:

A PDF version of the documentation is available here.

Syntax and Examples

Functions are called using an R-like syntax. Some general rules:

• Density functions: pystats.d*. For example, the Normal (Gaussian) density is called using

pystats.dnorm(<value>, <mean>, <standard deviation>)

• Cumulative distribution functions: pystats.p*. For example, the Gamma CDF is called using

pystats.pgamma(<value>, <shape parameter>, <scale parameter>)

• Quantile functions: pystats.q*. For example, the Beta quantile is called using

pystats.qbeta(<value>, <a parameter>, <b parameter>)

• Random sampling: pystats.r*. For example, to generate a single draw from the Logistic distribution:

pystats.rlogis(<location parameter>, <scale parameter>)

The library also supports lists as input/output:

• The pdf, cdf and quantile functions can take list arguments. For example:

norm_pdf_vals = pystats.dnorm([x / 10 for x in range(-10, 10, 1)], 1.0, 2.0)

• The randomization functions (r*) can output lists of arbitrary size. For example, the following code will generate a 100-item list of iid draws from a Gamma(3,2) distribution:

gamma_rvs = pystats.rgamma(100, 3.0, 2.0)

Additionally, most parameters have defaults to most common values and named parameters are also supported. For example, to generate a single draw from a Normal(0, 2) the following can be used:

norm_draw = pystats.rnorm(sd=2.0)

Examples

More examples with code:

# Evaluate the normal PDF at x = 1, mu = 0, sigma = 1
dval_1 = pystats.dnorm(1.0, 0.0, 1.0)
 
# Evaluate the normal PDF at x = 1, mu = 0, sigma = 1, and return the log value
dval_2 = pystats.dnorm(1.0, 0.0, 1.0, True)
 
# Same as above, but using default values and named parameters
dval_3 = pystats.dnorm(1.0, log=True)

# Evaluate the normal CDF at x = 1, mu = 0, sigma = 1
pval = pystats.pnorm(1.0, 0.0, 1.0)
 
# Evaluate the Laplacian quantile at q = 0.1, mu = 0, sigma = 1
qval = pystats.qlaplace(0.1, 0.0, 1.0)

# Draw from a t-distribution with dof = 30
rval = pystats.rt(dof=30)

# List output
beta_rvs = pystats.rbeta(100, 3.0, 2.0)

# List input
beta_cdf_vals = pystats.pbeta(beta_rvs, 3.0, 2.0)

For more information on default values, parameter names and other examples, check the documentation.

Credits

This library uses kthohr/stats for the statistical distribution functions, kthohr/gcem (a dependency of the previous library) and pybind/pybind11 to generate the binding code.

License

Copyright (c) Marc Izquierdo 2021
This library is licensed under the MIT License. See LICENSE for more details.