Introduction to Python - CONDA common commands

• Little knowledge , Great challenge ！ This article is participating in “ A programmer must have a little knowledge ” Creative activities .

This paper introduces the use of conda management anaconda Python Environment related commands .

conda Environment related commands

Create an environment

conda create -n env_name python=3.7 --clone another_env
 Copy code 

-n：name Indicates the name of the new environment

python： Use python edition

--clone： Copy from existing environment

Delete environment

conda remove -n env_name --all
 Copy code 

Look at the environment

conda env list 
 Copy code 

or

conda info -e
 Copy code 

Activate the environment

conda activate env_name
 Copy code 

or

source activate env_name
 Copy code 

Out of the environment

conda deactivate
 Copy code 

or

source deactivate
 Copy code 

Will return to base Environmental Science

conda Package related commands

View the current environment conda Managed python Package list

conda list
 Copy code 

install python package

conda install package_name # Installation package 
conda install package_name_1 package_name_2 package_name_3 ... # Install multiple packages at a time 
conda install package_name=1.1.0 # Install the specified version of the package 
 Copy code 

Update package

conda update package_name #  Update package 
conda upgrade --all # Update all packages 
 Copy code 

Uninstall package

conda remove package_name
 Copy code 

Search for packages with unclear names

conda search search_term
 Copy code 

conda Recreate the environment

Use conda management python An important consideration is portability ,conda Several methods are provided for reproducing a conda Environmental Science .

Clone

This command has been described above , Used to reproduce an environment locally

conda create --name new_env --clone old_env
 Copy code 

Spec List

Copy the environment between computers with the same operating system , Can generate spec list

#  Generate  spec list  file 
conda list --explicit > spec-list.txt 
# Recreate the environment ：
conda create --name python-course --file spec-list.txt
 Copy code 

Environment.yml

Use -export Option to generate a environment.yml file , To replicate the project environment between different platforms and operating systems .

spec list Document and environment.yml The difference between files is ： environment.yml Files are not specific to a particular operating system , And use YAML Format .environment.yml Only the package name is listed , from conda Build the environment based on the name of the package . Another difference is -export It also includes the use of pip Installed packages , and spec list There is no .

# export  environment.yml  file ：
conda env export > environment.yml
# Recreate the environment ：
conda env create -f environment.yml
 Copy code 

Be careful ： If the current path already has environment.yml file ,conda Will rewrite this file

Conda Pack

The above two reproduction methods are based on recording the current environment package information , To the idea of new machine reconstruction . and Conda Pack It is used to package the files of the current environment directly , The idea of unpacking the new machine .

conda-pack Specify the platform and operating system , The target computer must have the same platform and operating system as the source computer .

install conda pack

# from conda
conda install -c conda-forge conda-pack
# from pip
pip install conda-pack
 Copy code 

Packaging environment

conda pack -n my_env
conda pack -n my_env -o out_name.tar.gz
 Copy code 

Recreate the environment

mkdir -p path_to_my_new_env #  Advice on anaconda Of envs In the folder 
tar -xzf my_env.tar.gz -C path_to_my_new_env #  Unzip the files in the package 
source path_to_my_new_env/bin/activate #  Activate the environment 
(my_env) $ python #  Let's take a look at  python  Then exit 
(my_env) $ conda-unpack #  Very important , Please don't ignore 
 Copy code