Conda Env YML Outdated and Suggestion for Version Control Addition

[Charithwij]

The current conda yml environment contains several python built-in packages (which should be removed from the yml file) or packages supported by conda (which do not require a pip installation).
Of the dependencies on the current list:

• python (conda supported)
• scanpy (conda supported)
• pip:
  • numpy (conda supported)
  • pandas (conda supported)
  • itertools (built in python package)
  • time (built in python package)
  • tqdm (conda supported)
  • plotly (conda supported)

Additionally, the yml file is missing ipython, which seen in the tutorial's import lines.
One last suggestion would be to implement strict package version control in the conda env itself prevent conda installing higher version and to avoid future issues that may arise with updates to these packages.

[adeslatt]

I think the yml file is no longer needed for the tutorial if we go ahead with the jupyter lab notebook. In my forked version: https://github.com/adeslatt/NSForest/blob/master/ipynb/Tutorial.ipynb

you can see the identical to readthedocs. Once the pull request is accepted - we can add these packages in the markdown to say to install them when making a clean environment.