Solving a pip issue using a range vs removing package range altogether
Question:
I’m getting a conflict with a pip install and I get this message:
To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict
So if I have, let’s say:
pandas==1.1.2
I could put:
pandas>=1.1.0,<1.1.4
But would using:
pandas
simply look through all possibilities and fix it itself?
I’m not sure If I’m being clear, but basically, if I don’t care about version number, can I just use no versions as a better solution, because it would implicitly go through the all the ranges?
Answers:
Yes, running the command
pip install pandas
will find the most recent version that can be used.
While this might fix your issue as @linger1109 suggested, it is highly recommended to specify a version number in your dependency manifest (e.g. requirements.txt
.
There is no guarantee that future versions of pandas are fully backwards compatible, meaning the next time you install your dependencies, pip might opt for a newer version of pandas, potentially breaking your code.
I’d suggest finding versions of dependencies that are compatible with each other and declaring them in your dependencies manifest.
I’m getting a conflict with a pip install and I get this message:
To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict
So if I have, let’s say:
pandas==1.1.2
I could put:
pandas>=1.1.0,<1.1.4
But would using:
pandas
simply look through all possibilities and fix it itself?
I’m not sure If I’m being clear, but basically, if I don’t care about version number, can I just use no versions as a better solution, because it would implicitly go through the all the ranges?
Yes, running the command
pip install pandas
will find the most recent version that can be used.
While this might fix your issue as @linger1109 suggested, it is highly recommended to specify a version number in your dependency manifest (e.g. requirements.txt
.
There is no guarantee that future versions of pandas are fully backwards compatible, meaning the next time you install your dependencies, pip might opt for a newer version of pandas, potentially breaking your code.
I’d suggest finding versions of dependencies that are compatible with each other and declaring them in your dependencies manifest.