Tell me, why I should use Anaconda instead of Python Virtual Environment?

Discussion Zone
, , , ,
1

By my experience, setting up a conda environment is as complicated as it can get and I have to remember certain things to be able to activate and deactivate the environment. Also, the environment has to be turned on for me to use it. On the other hand, when I set up python virtual environment, I don’t always have to activate or deactivate depending on my situation. The most beautifull thing about python virtual environment is that, I can include the softfware as component of the data that I am processing and anyone would be able to run or reproduce my work without struggling to set up the python virtual environment. MHO.

So, tell me, why do you use anaconda instead of python virtual environment?

2

Conda environments can handle software beyond Python packages, so a conda environment can have non-Python libraries (like the Intel MKL) installed, compilers, the CUDA SDK, and so on. The dependency checking is also more thorough compared with pip. The drawbacks are more complexity as you said and you need the conda tool available to activate/deactivate/create conda environments.