scipy 1.13 breaks lapy

[m-reuter]

The behaviour of sparse matrix (CSC, CSR) creation has changed in script 1.13 breaking our code. Multiple data entries with the same indices are not summed up any longer but produce duplicate entries, as we do in our adjacency matrices.

I am fixing scipy to <= 1.12 for now in #60

For making lapy work with scipy 1.13 we should wait what they say about my report here:
scipy/scipy#20670

If this is not a bug, but a feature of 1.13, we can create a workaround by initialising the sparse matrix in COO format and then converting to CSC as that sums up duplicate entries again:
https://docs.scipy.org/doc/scipy/reference/sparse.html#example-2