Download Anaconda

Anaconda itself is an open source and free software. When downloading, you can directly search anaconda, visit the official website and enter the optional version interface

Pull to the bottom of the interface to select the version you want to download according to your system

After downloading, double-click the exe file to install it.
Note: try not to install it in Disk c, but in other larger disks.
During installation, please check "agree to add environment variables", which is useful later!!!!

Open the Anaconda command line after installation

Enter the instruction to create an environment name to be built in the hall

In the hall, Anaconda is not allowed to create an environment name with the same name, so it can build an independent small room to store its own environment

The code for creating a room is shown below.

// Create a separate room
conda create -n pytorch python=3.7
//Create means create
//-n means name
//Pytorch means the name of the created room. Here I use pytorch,
//You can also use a name you are familiar with and easy to remember
//python=3.7 means that I want to preset a bottom Python 3 environment in this room

Enter the room from the hall

It can be seen that anaconda will not automatically enter the created room and is still in the base environment. The previous picture also shows how to enter the created room.

// Enter room (active)
conda activate pytorch
// Exit the room
conda deactivate

View the list of bags already packed in the room

After entering the room, we can check which bags are in the room

conda list

There is no pytorch yet

Install pytorch into the room

pytorch, torch also means torch. It's not very different from tensorflow. It depends on the individual

If the above command line is copied directly, it will be downloaded through the pytorch official website by default, but it may take a long time for foreign servers to be installed.

The following describes how to modify Anaconda's default download source. Through Tsinghua source or university of science and technology source.

//Tsinghua source:
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/conda-forge
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/msys2/
conda config --set show_channel_urls true

//China University of science and technology source:
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/conda-forge/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/msys2/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/bioconda/
conda config --add channels https://mirrors.ustc.edu.cn/anaconda/cloud/menpo/
conda config --set show_channel_urls yes

Tsinghua source. After copying, click enter and then press enter to add it

China University of science and technology source, click enter after copying, and then press enter to add it

See which source is preferred or which sources have been added

// View source
conda config --show-source

// Install pytorch using domestic source, or copy the code just on the official website, but delete - c and the following contents
conda install pytorch torchvision torchaudio cpuonly

There are some endogenous in package countries that have not been updated

Use the address of the original official website

// Less common
conda install pytorch torchvision torchaudio cpuonly -c pytorch

Enter y to start downloading. If the network is unstable and some packages fail to download, you can execute the above instructions again for downloading and installation. It may take a long time here.

See which rooms are in the base environment

// Enter the following command to exit to the base Hall
conda deactivate
// See which rooms are available
conda info -e
// Select the room you want to enter
conda activate pytorch

Verify that pytorch is installed in the room

Execute python
Import torch
Success indicates that pytorch has been installed

Delete room in lobby

// Return to lobby delete
conda remove -n XXX --all
// XXX indicates the name of the room to be deleted
// --all delete everything in it

Import the newly established environment into pycharm

Similarly, download the software on pycharm's official website and open it after successful installation. After creating a new project, change its compiler to the python 3 interpreter in Python just established in Anaconda.
Click settings

Click on the gear

Click ADD

Select conda to create a new one or select an existing one

After modification, you can see

You can install the desired package to the room at the terminal
conda install XXX

Jupiter notebook editor

This is Anaconda's own editor. It is opened in the form of a web page. You can click to open it, as shown in the figure

Jupyter is an interactive editor that can display the results in real time. Pychart needs to be put together before it can be displayed

Another Jupiter notebook needs to be installed in the room

// Installing in pytorch
conda install jupyter notebook

The above operations refer to the video of the UP Master of station b, which can be viewed together with the video. The links are as follows

Installation video: Video Explanation.