Umap generation failed for tanimoto count metric

[julianpollmann]

Umap generation with CPU for tanimoto_distance_unfolded_count fails:

data_plot = create_chem_space_umap(
    data_compounds,  # dataframe with smiles and class/subclass etc. information
    col_smiles="smiles",
    inplace=False,
    x_col="x",
    y_col="y",
    fpgen = rdFingerprintGenerator.GetMorganGenerator(radius=9, fpSize=4096),
)

Stacktrace:

UMAP.fit(self, X, y, ensure_all_finite, **kwargs)
   2554     _m = self.metric if self._sparse_data else self._input_distance_func
-> 2555     dmat = pairwise_distances(X[index], metric=_m, **self._metric_kwds)
   2556 except (ValueError, TypeError) as e:
   2557     # metric is numba.jit'd or not supported by sklearn,
   2558     # fallback to pairwise special
...
TypeError: not enough arguments: expected 4, got 2

During handling of the above exception, another exception occurred:
...
TypingError: Failed in nopython mode pipeline (step: nopython frontend)
non-precise type pyobject
During: typing of argument at /xxx/chemap/lib/python3.13/site-packages/umap/distances.py (1283)

File "xxx/chemap/lib/python3.13/site-packages/umap/distances.py", line 1283:
def pairwise_special_metric(
    <source elided>

        @numba.njit(fastmath=True)
        ^

During: Pass nopython_type_inference 

This error may have been caused by the following argument(s):
- argument 0: Cannot determine Numba type of <class 'scipy.sparse._csr.csr_matrix'>
- argument 1: Cannot determine Numba type of <class 'scipy.sparse._csr.csr_matrix'>

[florian-huber]

This should be fixed with recent fix (wrong metrics were picked) --> 0.3.2