Overview and installation of scientific computing toolkit scipyscipy

Scipy Introduce

scipy It's a python Open source mathematical calculation library , Can be applied to Mathematics 、 Science and Engineering , It is based on numpy Scientific Computing Library of .

Mainly consists of statistical 、 optimization 、 linear algebra 、 integral 、 The Fourier transform 、 signal processing and The image processing as well as Ordinary differential equations And other calculations used in science and engineering

scipy Module introduction

scipy Mainly through the following packages Mathematical algorithm and Scientific Computing , The back for scipy The explanation of is mainly based on these packages

• cluster： Inclusion clustering algorithm
• constants： Some constants in physics and Mathematics
• fftpack： The fast Fourier transform
• integrate： Solution of integral and ordinary differential equations
• interpolate： Interpolation and smoothing spline function
• io： Input and output
• linalg： linear algebra
• ndimage：N Dimensional image processing
• ord： Regression orthogonal distance
• optimize： Optimization and root seeking equations
• signal： signal processing
• sparse： sparse matrix
• spatial： Spatial data structures and algorithms
• special： Special functions
• stats： Statistical distribution and function

scipy Related information

• scipy Official website ：https://scipy.org
• scipy file ：https://docs.scipy.org/
• scipy Source code ：https://github.com/scipy/scipy
• scipy Chinese Course ：https://wizardforcel.gitbooks.io/scipy-lecture-notes/content/

Reading guide

Last article Python Mathematical calculation library of scipy Introduce Has been described in detail in scipy Use of , This article mainly introduces how to install scipy, This article installs Python The package approach applies to other packages

install Python There are usually three ways to package ：

1. adopt pip To install
2. adopt conda To install
3. adopt whl File to install

Now for scipy The installation is also based on the above three main methods

pip Installation package

Python In fact, it is very convenient to install libraries in , adopt pip It can be easily installed

pip install scipy

Sometimes through pip Installing packages from the default source will be very slow , At this point we can go through modify pip Source To provide the speed of downloading packages , Yes Tsinghua source 、 Aliyuan 、 Douban source Etc. can be used , What I use here is Tsinghua source The setting method is as follows ：

• Temporary settings
  About the command window, it will switch to the default pip Source

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple some-package

• Set to default
  Permanent validity

python -m pip install --upgrade pippip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple

adopt conda To install

adopt conda It is very convenient to manage different projects in a virtual environment , In this way, there will be no conflict between different versions of the same package between different projects , Therefore, it is suggested to adopt different virtual environments for different projects .

• install anaconda
  If you want to by conda To manage the virtual environment, we need to install anaconda perhaps miniconda, The difference between the two is that anaconda After the installation, it will take up a lot of space , Because it also contains many other libraries . and miniconda It will take up much less space , It will be installed later conda My orders and Python Some basic packages .
  anaconda Installation package download address ：https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/
  miniconda Installation package download address ：https://mirrors.tuna.tsinghua.edu.cn/anaconda/miniconda/
  Download according to Different systems And what is needed Python edition Just choose

• Creating a virtual environment
  adopt conda create To create a virtual environment -n Used to set the name of the virtual environment (py3.8), hinder python== Virtual environment is used to create Python Version of

conda create -n py3.8 python==3.8

• Enter the virtual environment

activate py3.8

• conda Installation package

conda install scipy

adopt whl File to install

whl The file is actually the file format compiled by the package , The students may have found it carefully pip install scipy The library was downloaded when it was whl File to install , Then why do you need to use it alone whl File to install the package ？

Download separately whl File to install , Because sometimes pip When installing the package , Download is not whl File instead Source code , After downloading the source code, you need to compile it locally into whl File and then install , however Sometimes, due to local environment problems, compilation errors will occur , Some mistakes are troublesome to solve , At this time, use whl File installation can avoid such problems .

whl File download website ：https://www.lfd.uci.edu/~gohlke/pythonlibs/

At this time, you can according to your local python The version and number of bits of the system whl file , take whl After the file is downloaded locally , Re pass python To install , The installation command is as follows

pip install scipy-1.8.0-cp38-cp38-win_amd64.whl