Python Library 学习 - Conda

Python Library 学习 - Conda

1. package 操作

- 查看已安装的packages

conda list

- 安装

conda install $PACKAGE_NAME
conda install -n $ENV_NAME $PACKAGE_NAME
# example
conda install numpy

批量安装

conda install --yes --file requirements.txt -c conda-forge

- 指定源安装

conda install --channel $URL $PACKAGE_NAME
conda install -c $URL $PACKAGE_NAME
# "--channel" can be shortened as "-c"
# we can aslo use $CHANNEL_NAME replace $URL
# example
conda install -c conda-forge numpy

- 卸载

conda remove $PACKAGE_NAME
conda remove --name $ENVIRONMENT_NAME $PACKAGE_NAME
conda remove -n $ENVIRONMENT_NAME $PACKAGE_NAME
# "--name" can be shortened as "-n"

- 更新

更新 package 可以使用 update 命令或者 upgrade 命令

conda update $PACKAGE_NAME
conda update --name $ENVIRONMENT_NAME $PACKAGE_NAME
# 更新所有 packages
conda update --all

conda upgrade $PACKAGE_NAME
conda upgrade --name $ENVIRONMENT_NAME $PACKAGE_NAME
# 更新所有 packages
conda upgrade --all

- 导出环境中所有 package

conda list -myenv > requirements.txt

2. channel 操作

2.1. 常用命令

- 显示所有channel

conda config --show channels

- 查看当前channel 的状态和优先级

conda config --get channels

- 添加channel，并设置为最高优先级

conda config --add channels $CHANNEL_NAME
conda config --add channels $CHANNEL_URL
# example
conda config --add channels conda-forge

- conda 默认源

conda config --remove channels $CHANNEL_NAME
conda config --remove channels $CHANNEL_URL

- 设置channel的严格优先

conda config --set channel_priority strict

2.2. 国内常用的镜像 channel

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
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/

3. env 操作: 虚拟环境

- 查看虚拟环境

conda env list 
conda info -e

- 创建虚拟环境

conda create --name $ENVIRONMENT_NAME python=X.X
conda create -n $ENVIRONMENT_NAME python=X.X
# "--name" can be shortened as "-n"
# example
conda create --name py38 python=3.8
conda create -n py38 python=3.8

- 激活虚拟环境

conda activate $ENVIRONMENT_NAME

- 退出虚拟环境，返回base环境

conda deactivate

- 删除虚拟环境

conda remove --name $ENVIRONMENT_NAME --all

4. solver

libmamba, 更快的计算器用于计算库之间的依赖

conda install -n $ENVIRONMENT_NAME conda-libmamba-solver
conda config --set solver libmamba

# 原始的 solver
conda config --set solver classic

5. 常见问题

1. 安装包不存在

conda install librosa

当出现以下错误提示时：

ackagesNotFoundError: The following packages are not available from current channels:

1. 执行anaconda search -t conda X，显示可用的版本

2. 选择适合自己的版本，执行：conda install -c https://conda.anaconda.org/X X为所选择的安装的包

   如：conda install -c https://conda.anaconda.org/conda-forge librosa

Reference

[1] Conda vs. pip vs. virtualenv commands, Conda documents

[2] A Faster Solver for Conda: Libmamba, 2022, site