This is not really related to the curriculum…
I’ve installed python3.9 yesterday on my windows 10…
Before installing I kept note of all the installed packages using pip freeze>req.txt.
Now after installing v3.9 I uninstalled python3.8.6 from control-panel>programs. it was all fine but I lost my installed packages.
So, today I was going to install all the packages with pip install -r req.txt but pip gave me error…
I fixed it by following folks on Stack Overflow, by deleting python38 directory…
But I’m unable to install them , I came across following error
Delete python
Download python from their website. One version old , just to show you the process
Then upgrade via the cmd line. Do not delete the previous version.
Upgrade pip as well
I asked u to install the previous version just to show u the process. U don’t need to do it. But u have to do it this time
Upgrade pip if there’s an update.
Next time u want to upgrade just
Run the command to upgrade python version.
If it prompts to update the pip version just do it.
Hi @mukeshgurpude,
It seems like it’s trying to use some cached version of your libraries. This issue is not new on Windows 10 [see this] (https://github.com/pypa/pip/issues/8272).
Since Python 3.9 is ery new, it may introduce conflicting behaviours or incompatibilities with some of the libraries you have installed in your system.
Before considering rolling back to a 3.8.x version, I would try to completely remove all trace of python by uninstalling from the Control Panel as you did, AND, physically deleting the folders; specially site-packages. This would permit you a clean install. the caveat is, if you have python programs you need for some other tasks, you will need to reinstall their dependencies too!
Cheers,
About upgrading python from the command line, there is no way on Windows to my knowledge if you use the standard installable from python.org. On a Mac, you can do this via homebrew. Maybe this is the way @gururajankappa refers to.
Actually, not only numpy there are many packages not optimized for python3.9…
Now I understood the importance of virtual environments for each project.
You can “watch” their Github. I can’t think of nothing better right now, sorry! But, this way, you’ll see on your Github “feed” when they release a new version.