Virtual Environment For Installing Tensorflow : Why Do I need it for Whiich Purpose?

Question

Following the thankful youtube links, just had succeeded easily to install tensorflow through my Anaconda Prompt.

What I had done is :

1) conda create -n tensor2 python=3

first I had created a virtual environment names tensor2(I don't know why do I need this) and also What does -n refer to?

2) then activate tensor2 Virtual Environment then run the pip install tensorflow.

So Wrapping up the question:

1) Why do I need to create Virtual Enviornmnet especially for module tensorflow while other modules are just using pip install instantly?

2) additional question is, what does -n refer to at above command? and also, what does -m refer to in the statement "python -m pip install /module name/'?

Answer 1

What does -n refer to in conda create

https://conda.io/docs/using/envs.html#create-an-environment --name or -n just defines what the environment will be named after.

What does -m refer to in python -m

The -m flag search for a given module (in your case it searches for the pip module) and if found, it runs the __main__ function (if you simply run python -m pip it will display the help output). And 'intall module' are then the arguments passed to pip. This is often done to circumvent that 'pip' would not be in your path but that python would be able to locate it (throught its python path).

Why do you need virtual environments ?

Well actually you don't. Why are people suggesting to use them ? Because it gives you some more isolated environments to experiment with, without 'damaging' the rest of your system, if you experiment a lot there is chance the some dependencies could go in conflict and it also allows you to switch between different versions. If you plan to only use one environment and keep it up to date (or freeze it to a certain version) then there's no need to.