Note from Lupin on install errors

[jkanner]

Tuto 1.4:

I obtained the following (error ??) message when I performed the installation of PYCBC 1.18.0 and LALsuit 6.82 although I can successfully run all the codes in each code cell. Should we take care it?
—

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
tensorflow 2.8.0 requires tf-estimator-nightly==2.8.0.dev2021122109, which is not installed.
tensorflow 2.8.0 requires numpy>=1.20, but you have numpy 1.19.5 which is incompatible.
datascience 0.10.6 requires folium==0.2.1, but you have folium 0.8.3 which is incompatible.
albumentations 0.1.12 requires imgaug<0.2.7,>=0.2.5, but you have imgaug 0.2.9 which is incompatible.

[jkanner]

@linlupin Thanks for flagging this.

I tried removing the version numbers for pyCBC and lalsuite, the tutorial 1.4 seems to run without error. I'm tempted to try removing the version numbers from all of the pip install commands. There are pros and cons to doing it this way, but in principle, always having the latest version of a given package should be an OK default.

Do you agree we should try removing version numbers from the pip install commands? If needed, we could always add back in a specific version for some tutorials.

[linlupin]

No problem. I suppose that we can just remove the version number. But it is best to test whether the installation of the latest version has any bug or not.