Installation instructions - add common error

[odunbar]

Issue

An error we have discovered is that sometimes Python is seemingly not linked properly when installing PyCall/Conda
But seems sporadic

Solution

In the installation instructions docs.
Add links to PyCall.jl, and Conda.jl documentation.
And section below the installation instructions with common errors. Some example text:

Sometimes, if encountering an error such as

ERROR: InitError: PyError (PyImport_ImportModule

The Python package sklearn.gaussian_process.kernels could not be imported by pyimport.

Often this can be fixed by first rebuilding PyCall

julia --project
julia> ENV["PYTHON"]=""
julia> Pkg.build("PyCall")
julia> exit()

then calling the Conda install packages e.g....

julia --project -e 'using Conda; Conda.add("scipy=1.8.1", channel="conda-forge")'

[odunbar]

Another Build error.

For python 3.12 one cannot build with scipy=1.8.1 and scikit-learn=1.1.1

These should be updated to their current versions (which appear to work fine: scipy = 1.14.1 , scikit-learn=1.5.1 (prefer to pin to new versions over leaving it free to update unit tests etc.)