Statistical functions (scipy.stats
)#
This module contains a large number of probability distributions, summary and frequency statistics, correlation functions and statistical tests, masked statistics, kernel density estimation, quasi-Monte Carlo functionality, and more.
Statistics is a very large area, and there are topics that are out of scope for SciPy and are covered by other packages. Some of the most important ones are:
statsmodels: regression, linear models, time series analysis, extensions to topics also covered by
scipy.stats
.Pandas: tabular data, time series functionality, interfaces to other statistical languages.
PyMC: Bayesian statistical modeling, probabilistic machine learning.
scikit-learn: classification, regression, model selection.
Seaborn: statistical data visualization.
rpy2: Python to R bridge.
Probability distributions#
Each univariate distribution is an instance of a subclass of rv_continuous
(rv_discrete
for discrete distributions):
|
A generic continuous random variable class meant for subclassing. |
|
A generic discrete random variable class meant for subclassing. |
|
Generates a distribution given by a histogram. |
Continuous distributions#
An alpha continuous random variable. |
|
An anglit continuous random variable. |
|
An arcsine continuous random variable. |
|
Argus distribution |
|
A beta continuous random variable. |
|
A beta prime continuous random variable. |
|
A Bradford continuous random variable. |
|
A Burr (Type III) continuous random variable. |
|
A Burr (Type XII) continuous random variable. |
|
A Cauchy continuous random variable. |
|
A chi continuous random variable. |
|
A chi-squared continuous random variable. |
|
A cosine continuous random variable. |
|
Crystalball distribution |
|
A double gamma continuous random variable. |
|
A double Weibull continuous random variable. |
|
An Erlang continuous random variable. |
|
An exponential continuous random variable. |
|
An exponentially modified Normal continuous random variable. |
|
An exponentiated Weibull continuous random variable. |
|
An exponential power continuous random variable. |
|
An F continuous random variable. |
|
A fatigue-life (Birnbaum-Saunders) continuous random variable. |
|
A Fisk continuous random variable. |
|
A folded Cauchy continuous random variable. |
|
A folded normal continuous random variable. |
|
A generalized logistic continuous random variable. |
|
A generalized normal continuous random variable. |
|
A generalized Pareto continuous random variable. |
|
A generalized exponential continuous random variable. |
|
A generalized extreme value continuous random variable. |
|
A Gauss hypergeometric continuous random variable. |
|
A gamma continuous random variable. |
|
A generalized gamma continuous random variable. |
|
A generalized half-logistic continuous random variable. |
|
A generalized hyperbolic continuous random variable. |
|
A Generalized Inverse Gaussian continuous random variable. |
|
A Gibrat continuous random variable. |
|
A Gompertz (or truncated Gumbel) continuous random variable. |
|
A right-skewed Gumbel continuous random variable. |
|
A left-skewed Gumbel continuous random variable. |
|
A Half-Cauchy continuous random variable. |
|
A half-logistic continuous random variable. |
|
A half-normal continuous random variable. |
|
The upper half of a generalized normal continuous random variable. |
|
A hyperbolic secant continuous random variable. |
|
An inverted gamma continuous random variable. |
|
An inverse Gaussian continuous random variable. |
|
An inverted Weibull continuous random variable. |
|
An Irwin-Hall (Uniform Sum) continuous random variable. |
|
Jones and Faddy skew-t distribution. |
|
A Johnson SB continuous random variable. |
|
A Johnson SU continuous random variable. |
|
Kappa 4 parameter distribution. |
|
Kappa 3 parameter distribution. |
|
Kolmogorov-Smirnov one-sided test statistic distribution. |
|
Kolmogorov-Smirnov two-sided test statistic distribution. |
|
Limiting distribution of scaled Kolmogorov-Smirnov two-sided test statistic. |
|
A Laplace continuous random variable. |
|
An asymmetric Laplace continuous random variable. |
|
A Levy continuous random variable. |
|
A left-skewed Levy continuous random variable. |
|
A Levy-stable continuous random variable. |
|
A logistic (or Sech-squared) continuous random variable. |
|
A log gamma continuous random variable. |
|
A log-Laplace continuous random variable. |
|
A lognormal continuous random variable. |
|
A loguniform or reciprocal continuous random variable. |
|
A Lomax (Pareto of the second kind) continuous random variable. |
|
A Maxwell continuous random variable. |
|
A Mielke Beta-Kappa / Dagum continuous random variable. |
|
A Moyal continuous random variable. |
|
A Nakagami continuous random variable. |
|
A non-central chi-squared continuous random variable. |
|
A non-central F distribution continuous random variable. |
|
A non-central Student's t continuous random variable. |
|
A normal continuous random variable. |
|
A Normal Inverse Gaussian continuous random variable. |
|
A Pareto continuous random variable. |
|
A pearson type III continuous random variable. |
|
A power-function continuous random variable. |
|
A power log-normal continuous random variable. |
|
A power normal continuous random variable. |
|
An R-distributed (symmetric beta) continuous random variable. |
|
A Rayleigh continuous random variable. |
|
A relativistic Breit-Wigner random variable. |
|
A Rice continuous random variable. |
|
A reciprocal inverse Gaussian continuous random variable. |
|
A semicircular continuous random variable. |
|
A skewed Cauchy random variable. |
|
A skew-normal random variable. |
|
A studentized range continuous random variable. |
|
A Student's t continuous random variable. |
|
A trapezoidal continuous random variable. |
|
A triangular continuous random variable. |
|
A truncated exponential continuous random variable. |
|
A truncated normal continuous random variable. |
|
An upper truncated Pareto continuous random variable. |
|
A doubly truncated Weibull minimum continuous random variable. |
|
A Tukey-Lamdba continuous random variable. |
|
A uniform continuous random variable. |
|
A Von Mises continuous random variable. |
|
A Von Mises continuous random variable. |
|
A Wald continuous random variable. |
|
Weibull minimum continuous random variable. |
|
Weibull maximum continuous random variable. |
|
A wrapped Cauchy continuous random variable. |
The fit
method of the univariate continuous distributions uses
maximum likelihood estimation to fit the distribution to a data set.
The fit
method can accept regular data or censored data.
Censored data is represented with instances of the CensoredData
class.
|
Instances of this class represent censored data. |
Multivariate distributions#
A multivariate normal random variable. |
|
A matrix normal random variable. |
|
A Dirichlet random variable. |
|
A Dirichlet multinomial random variable. |
|
A Wishart random variable. |
|
An inverse Wishart random variable. |
|
A multinomial random variable. |
|
A Special Orthogonal matrix (SO(N)) random variable. |
|
An Orthogonal matrix (O(N)) random variable. |
|
A matrix-valued U(N) random variable. |
|
A random correlation matrix. |
|
A multivariate t-distributed random variable. |
|
A multivariate hypergeometric random variable. |
|
Contingency tables from independent samples with fixed marginal sums. |
|
A vector-valued uniform direction. |
|
A von Mises-Fisher variable. |
scipy.stats.multivariate_normal
methods accept instances
of the following class to represent the covariance.
Representation of a covariance matrix |
Discrete distributions#
A Bernoulli discrete random variable. |
|
A beta-binomial discrete random variable. |
|
A beta-negative-binomial discrete random variable. |
|
A binomial discrete random variable. |
|
A Boltzmann (Truncated Discrete Exponential) random variable. |
|
A Laplacian discrete random variable. |
|
A geometric discrete random variable. |
|
A hypergeometric discrete random variable. |
|
A Logarithmic (Log-Series, Series) discrete random variable. |
|
A negative binomial discrete random variable. |
|
A Fisher's noncentral hypergeometric discrete random variable. |
|
A Wallenius' noncentral hypergeometric discrete random variable. |
|
A negative hypergeometric discrete random variable. |
|
A Planck discrete exponential random variable. |
|
A Poisson discrete random variable. |
|
A uniform discrete random variable. |
|
A Skellam discrete random variable. |
|
A Yule-Simon discrete random variable. |
|
A Zipf (Zeta) discrete random variable. |
|
A Zipfian discrete random variable. |
An overview of statistical functions is given below. Many of these functions
have a similar version in scipy.stats.mstats
which work for masked arrays.
Summary statistics#
|
Compute several descriptive statistics of the passed array. |
|
Compute the weighted geometric mean along the specified axis. |
|
Calculate the weighted harmonic mean along the specified axis. |
|
Calculate the weighted power mean along the specified axis. |
|
Compute the kurtosis (Fisher or Pearson) of a dataset. |
|
Return an array of the modal (most common) value in the passed array. |
|
Calculate the nth moment about the mean for a sample. |
|
Compute the expectile at the specified level. |
|
Compute the sample skewness of a data set. |
|
Return the n th k-statistic ( |
|
Return an unbiased estimator of the variance of the k-statistic. |
|
Compute the trimmed mean. |
|
Compute the trimmed variance. |
|
Compute the trimmed minimum. |
|
Compute the trimmed maximum. |
|
Compute the trimmed sample standard deviation. |
|
Compute the trimmed standard error of the mean. |
|
Compute the coefficient of variation. |
|
Find repeats and repeat counts. |
|
Assign ranks to data, dealing with ties appropriately. |
|
Tie correction factor for Mann-Whitney U and Kruskal-Wallis H tests. |
|
Return mean of array after trimming a specified fraction of extreme values |
|
Calculate the geometric standard deviation of an array. |
|
Compute the interquartile range of the data along the specified axis. |
|
Compute standard error of the mean. |
|
Bayesian confidence intervals for the mean, var, and std. |
|
'Frozen' distributions for mean, variance, and standard deviation of data. |
|
Calculate the Shannon entropy/relative entropy of given distribution(s). |
|
Given a sample of a distribution, estimate the differential entropy. |
|
Compute the median absolute deviation of the data along the given axis. |
Frequency statistics#
|
Return a cumulative frequency histogram, using the histogram function. |
|
Compute the percentile rank of a score relative to a list of scores. |
|
Calculate the score at a given percentile of the input sequence. |
|
Return a relative frequency histogram, using the histogram function. |
|
Compute a binned statistic for one or more sets of data. |
|
Compute a bidimensional binned statistic for one or more sets of data. |
|
Compute a multidimensional binned statistic for a set of data. |
Quasi-Monte Carlo#
Contingency Tables#
Masked statistics functions#
- Statistical functions for masked arrays (
scipy.stats.mstats
)- Summary statistics
- Frequency statistics
- Correlation functions
- Statistical tests
- Transformations
- Other
Other statistical functionality#
Transformations#
|
Return a dataset transformed by a Box-Cox power transformation. |
|
Compute optimal Box-Cox transform parameter for input data. |
|
The boxcox log-likelihood function. |
|
Return a dataset transformed by a Yeo-Johnson power transformation. |
|
Compute optimal Yeo-Johnson transform parameter. |
|
The yeojohnson log-likelihood function. |
|
Compute the O'Brien transform on input data (any number of arrays). |
|
Perform iterative sigma-clipping of array elements. |
|
Slice off a proportion of items from both ends of an array. |
|
Slice off a proportion from ONE end of the passed array distribution. |
|
Calculate the relative z-scores. |
|
Compute the z score. |
|
Compute the geometric standard score. |
Statistical distances#
|
Compute the Wasserstein-1 distance between two 1D discrete distributions. |
|
Compute the Wasserstein-1 distance between two N-D discrete distributions. |
|
Compute the energy distance between two 1D distributions. |
Sampling#
Random variate generation / CDF Inversion#
|
Generate random samples from a probability density function using the ratio-of-uniforms method. |
Fitting / Survival Analysis#
Directional statistical functions#
|
Computes sample statistics for directional data. |
|
Compute the circular mean of a sample of angle observations. |
|
Compute the circular variance of a sample of angle observations. |
|
Compute the circular standard deviation of a sample of angle observations. |
Sensitivity Analysis#
|
Global sensitivity indices of Sobol'. |
Plot-tests#
|
Calculate the shape parameter that maximizes the PPCC. |
|
Calculate and optionally plot probability plot correlation coefficient. |
|
Calculate quantiles for a probability plot, and optionally show the plot. |
|
Compute parameters for a Box-Cox normality plot, optionally show it. |
|
Compute parameters for a Yeo-Johnson normality plot, optionally show it. |
Univariate and multivariate kernel density estimation#
|
Representation of a kernel-density estimate using Gaussian kernels. |
Warnings / Errors used in scipy.stats
#
|
Warns when data is degenerate and results may not be reliable. |
|
Warns when all values in data are exactly equal. |
|
Warns when all values in data are nearly equal. |
|
Represents an error condition when fitting a distribution to data. |
Result classes used in scipy.stats
#
Warning
These classes are private, but they are included here because instances of them are returned by other statistical functions. User import and instantiation is not supported.