Clustering graph. Python

Question

There is a database that has the following structure:

"vertex1" "vertex2" weight

For example:

"Word1" "Word2" 0.3

"Slovo3" "Slovo1" 0.8

"Word4" "Word2" 0.001

It is necessary to create a graph from such a base and then cluster it. Unfortunately, I did not understand the graph_tool how to apply (the documentation is scarce + it is not clear how to fasten the edge weight to clustering).

And installing SciPy on my windows 10 turned out to be a living hell, which I could not overcome.

Maybe there are more tools for solving this problem, or should we approach it in a different way?

I recommend asking a separate question "how to install scipy" on Windows 10 (in fact, scipy is a distribution that disguises itself as a Python module, precisely to facilitate the installation of scientific packages before the existence of a conda, support binary wheels in pip, etc.).
A typical installation process does not require any compilation.
Before moving on, learn how to install Python modules (you should not create your own library for working with graphs, just because you did not manage to install a package).
that is why I asked them not to write bicycles (for a long time it would be, and the terms are not rubber).
Hike 64-bit versions of python modules on Windows cannot be installed (because they are not installed).
this is not an answer (it does not say anything about clustering), it is a comment (in any case, the form of the answer is below and not here).
Once again: if you don’t know how to set up scipy (64-bit or not, satisfying some unnamed property or not), then ask a separate detailed (with all requirements) question (tell us what specific steps you performed, what you expected to get, instead of this is obtained in detail, step by step).
I just went to scipy.org and pressed the install button that leads to a page with instructions that look reasonable (make sure they don’t work for you)

Clustering graph. Python

0

More articles: