build: switch to conda dssp installation

[gcroci2]

With this PR, dssp package is installed through conda instead of source or apt-get. This is a usage improvement for the users and is a step forward for deploying deeprank2 on conda (see issue #559). In particular:

• now the appropriate conda channels (pointing to the right packages' releases) are specified in the action file
• dssp (version 4.2.2.1) is installed via conda
• this solves Fix Coveralls coverage 0% #542, by placing an init file into the tests folder. This is solved by ci: implement ruff #554 as well since Ruff automatically added the init file to the tests folder (and each subfolder, even if is not necessary for coveralls to detect all the files). I also added a check in the coveralls workflow file to make the action fail if the coverage drops below 80%.
• just to be sure since we do not have an action for testing the dockerfile, be sure to run that procedure again

[DaniBodor]

Looks good! I tested all 3 installation procedures (docker, env file, manual) and they all work!

I added 2 commits to this branch.

1. The first one is some automatic formatting updates. I think maybe you don't have the prettier extension installed, so then vs code cannot automatically format the files that use that (most non-python files).
2. Suggestions to the README.
   I thought it'd be easier to just put everything in a single commit, rather than add a bunch of code suggestions. Feel free to drop either if you don't like it or to adjust as you prefer.

Note that the python 3.11 issue isn't (stably) fixed after all, so please do not close that issue with this PR.

Finally, the newest version of pandas is giving a warning. Maybe we should add Pyarrow to the deeprank dependencies to make it future proof (and avoid the warning).

/tmp/ipykernel_32/1088082951.py:2: DeprecationWarning:
Pyarrow will become a required dependency of pandas in the next major release of pandas (pandas 3.0),
(to allow more performant data types, such as the Arrow string type, and better interoperability with other libraries)
but was not found to be installed on your system.
If this would cause problems for you,
please provide us feedback at pandas-dev/pandas#54466

import pandas as pd

[gcroci2]

Looks good! I tested all 3 installation procedures (docker, env file, manual) and they all work!

I added 2 commits to this branch.

1. The first one is some automatic formatting updates. I think maybe you don't have the prettier extension installed, so then vs code cannot automatically format the files that use that (most non-python files).

Oh right, I installed it now :)

2. Suggestions to the README.
   I thought it'd be easier to just put everything in a single commit, rather than add a bunch of code suggestions. Feel free to drop either if you don't like it or to adjust as you prefer.

Very nice edits, thanks! I'll leave the commit, I only corrected two typos and edited a couple of subtitles.

Note that the python 3.11 issue isn't (stably) fixed after all, so please do not close that issue with this PR.

I removed it from my initial comment indeed. What if I remove Python 3.11 from the CI for now? In a few weeks we can try to pick up the issue again, check it better, and see if anything has changed. But at least we will publish a "real" stable version that works on Python 3.10. I have added a commit for this, let me know what you think. If you agree, I will also add a comment to issue #540. @DaniBodor

Finally, the newest version of pandas is giving a warning. Maybe we should add Pyarrow to the deeprank dependencies to make it future proof (and avoid the warning).

I think they will add it themselves when installing pandas (as a pandas dependency in their config/toml file). The warning should be to gather feedback on this possibility shortly so that they know if adding this dependency will cause an issue for anyone. But I wouldn't worry about it, it will be installed automatically by pandas when the time comes.

[DaniBodor]

I noticed an indent level error in the installation.md file (which I probably created when copy/pasting from the README).

[DaniBodor]

Looks good! I tested all 3 installation procedures (docker, env file, manual) and they all work!
I added 2 commits to this branch.

1. The first one is some automatic formatting updates. I think maybe you don't have the prettier extension installed, so then vs code cannot automatically format the files that use that (most non-python files).

Oh right, I installed it now :)

2. Suggestions to the README.
   I thought it'd be easier to just put everything in a single commit, rather than add a bunch of code suggestions. Feel free to drop either if you don't like it or to adjust as you prefer.

Very nice edits, thanks! I'll leave the commit, I only corrected two typos and edited a couple of subtitles.

Note that the python 3.11 issue isn't (stably) fixed after all, so please do not close that issue with this PR.

I removed it from my initial comment indeed. What if I remove Python 3.11 from the CI for now? In a few weeks we can try to pick up the issue again, check it better, and see if anything has changed. But at least we will publish a "real" stable version that works on Python 3.10. I have added a commit for this, let me know what you think. If you agree, I will also add a comment to issue #540. @DaniBodor

Finally, the newest version of pandas is giving a warning. Maybe we should add Pyarrow to the deeprank dependencies to make it future proof (and avoid the warning).

I think they will add it themselves when installing pandas (as a pandas dependency in their config/toml file). The warning should be to gather feedback on this possibility shortly so that they know if adding this dependency will cause an issue for anyone. But I wouldn't worry about it, it will be installed automatically by pandas when the time comes.

Perfect. I agree with all of this.