Mamba is here, Python conda accelerates magically, and experiences the feeling of flying

If you use conda to download the Python package and build the environment on weekdays, and you clearly set the domestic source, it is still very slow, very slow, very slow,

Are you crazy, have a similar experience, like or watch at the end of the article.

This issue introduces a Python package manager Mamba:

• Use multi-threaded downloadrepository data and packages;
• Mamba uses libsolvspeeds up the resolution of dependencies (reduces the time of Solving environment: circles in the picture above);
• The core part of Mamba is implemented in C++ to achieve maximum execution efficiency;
• The most important thing is that Mamba is perfectly compatible with conda (replace conda in conda execution command with mamba).

mamba installation

Install directly with conda, the following line solves,

 conda install mamba -n base -c conda-forge

If conda is not configured, you can simply configure it by referring to the previous article:

Detailed Python CONDA

Using Anaconda to build a Python environment under Windows and Linux

Simply compare the difference between mamba and conda,

Conda on the left, mamba on the right, almost exactly the same, so after using conda, mamba can be used with almost no learning.

mamba installation package

#Take plotnine installation as an examplemamba install plotnine #Perfect replacement for turtle speed conda install plotnine

3.00 MB/s is quite gratifying, after all conda is often several kb/s!

mamba creates a Python environment

mamba create --name python3.5 python=3.5 #Perfect replacement for conda

The above interface appears, the build is successful, here Note️, the activation and exit of the environment can only be done using conda activate and conda deactivate

conda activate python3.5 #Activate python3.5 environmentconda deactivate #Exit python3.5 environment

Further study:

https://github.com/mamba-org/mamba