Select the best Python interpreter for me via machine learning.

[unikcc]

I have many virtual environments built via conda on my computer, and each time I start a new project, I have to select a correct Python Interpreter manually.
Although I can set default interpreter in setting, but it's not the best choice. Because I often create new environment and new project for specific purpose or switch environment betwen different project.
Usually, environment that satisfies such conditions are more likely selected, which are 1. I use most frequently recently. 2. Created just before a moment current project is created 3. Used by last opened project.
I think the Python Interpreter selection data can be recorded, feature dimension can contains items list above and more(e.g. string-level similarity between project name and environment name). Then train a SVM or anyother lightweight machine learning algorithm to recommend the best or topk python Interpreter when a new project is created.

[brettcannon]

Thanks for the idea, but we would rather not collect this amount of potentially personally-identifiable information in order to train a model. We also don't think the burden of selecting an environment once per workspace is burdensome enough to warrant the cost of maintaining the model. As well, we do not condone reusing environments and instead suggest users create environments per workspace.

Finally, #16994 will allow someone to implement this feature request themselves if they feel motivated enough to do so.