统计处理包Statsmodels: statistics in python

亦凉 2022-09-25 08:18 148阅读 0赞

http://[blog.csdn.net/pipisorry/article/details/52227580][blog.csdn.net_pipisorry_article_details_52227580]

## Statsmodels    ##

Statsmodels is a Python package that provides a complement to scipy for statistical computations including descriptive statistics and estimation of statistical models.

statsmodels原名叫scikits.statsmodels，已经改成了statsmodels。

其中也有时间序列分析的模块\[[Time Series analysis tsa][]\]

## statsmodels主要功能features ##

regression: Generalized least squares (including weighted least squares and least squares with autoregressive errors), ordinary least squares.  
    glm: Generalized linear models with support for all of the one-parameter exponential family distributions.  
    discrete choice models: Poisson, probit, logit, multinomial logit  
    rlm: Robust linear models with support for several M-estimators.  
    tsa: Time series analysis models, including ARMA, AR, VAR  
    nonparametric : (Univariate) kernel density estimators  
    datasets: Datasets to be distributed and used for examples and in testing.  
    PyDTA: Tools for reading Stata .dta files into numpy arrays.  
    stats: a wide range of statistical tests  
    **sandbox**: There is also a sandbox which contains code for generalized additive models (untested), mixed effects models, cox proportional hazards model (both are untested and still dependent on the nipy formula framework), generating descriptive statistics, and printing table output to ascii, latex, and html. There is also experimental code for systems of equations regression, time series models, panel data estimators and information theoretic measures. None of this code is considered "production ready".

### 安装 ###

pip install statsmodels

需要的依赖有：

Python >= 2.6, including Python 3.x

NumPy >= 1.5.1

SciPy >= 0.9.0

Pandas >= 0.7.1

Patsy >= 0.3.0

Cython >= 20.1, Needed if you want to build the code from github and not a source distribution. You must use Cython >= 0.20.1 if you’re on Python 3.4. Earlier versions may work for Python < 3.4.

**如果安装不上就源码安装：**

git clone git://github.com/statsmodels/statsmodels.git

cd statsmodels

python setup.py install

virtualenv中安装出错

error: Command "x86\_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -D\_FORTIFY\_SOURCE=2 -fPIC -I/home/piting/ENV/ubuntu\_env/lib/python3.4/site-packages/numpy/core/include -I/usr/include/python3.4m -I/home/piting/ENV/ubuntu\_env/include/python3.4m -c statsmodels/nonparametric/\_smoothers\_lowess.c -o build/temp.linux-x86\_64-3.4/statsmodels/nonparametric/\_smoothers\_lowess.o" failed with exit status 1

需要安装sudo apt-get install python3-dev\[[python-dev][]\]

python-dev: Header files and a static library for Python

\[[Installation][]\][  
][Link 1]

[皮皮blog][Link 1]

# statsmodels的使用 #

## 熵计算模块    ##

如renyi熵的计算

from statsmodels.sandbox.infotheo import renyientropy
     renyientropy(ij[np.nonzero(ij)] / sum(ij), alpha=q, logbase=math.e)

from: [http://blog.csdn.net/pipisorry/article/details/52227580][blog.csdn.net_pipisorry_article_details_52227580]

ref: \[[statsmodels][]/[statsmodels][statsmodels 1] github\]

\[homepage [Statsmodels][]\]

\[[Statsmodels’s Documentation][Statsmodels_s Documentation]\]

\[[statsmodels源码][statsmodels 2]\]

[blog.csdn.net_pipisorry_article_details_52227580]: http://blog.csdn.net/pipisorry/article/details/52227580
[Time Series analysis tsa]: http://statsmodels.sourceforge.net/stable/tsa.html
[python-dev]: https://sourceforge.net/projects/tdm-gcc/files/Sources/Vanilla%20Sources/_Deprecated/python-dev/
[Installation]: http://statsmodels.sourceforge.net/stable/install.html
[Link 1]: http://blog.csdn.net/pipisorry
[statsmodels]: https://github.com/statsmodels
[statsmodels 1]: https://github.com/statsmodels/statsmodels
[Statsmodels]: http://statsmodels.sourceforge.net/
[Statsmodels_s Documentation]: http://statsmodels.sourceforge.net/stable/index.html
[statsmodels 2]: http://pydoc.net/Python/scikits.statsmodels/0.3.1/