fda_kmeans.fit_predict(X)

[jack-pan-ai]

Bug description summary

---

TypeError Traceback (most recent call last)
/var/folders/js/lpf4z00557sb1m38br8bzj_476fsrl/T/ipykernel_16178/3403943172.py in
8 random_state=0,
9 )
---> 10 fda_clusters = fda_kmeans.fit_predict(X)

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/skfda/_utils/_sklearn_adapter.py in fit_predict(self, X, y)
157 y: object = None,
158 ) -> NDArrayInt:
--> 159 return super().fit_predict(X, y) # type: ignore[no-any-return]
160
161

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/sklearn/base.py in fit_predict(self, X, y)
749 # non-optimized default implementation; override when a better
750 # method is possible for a given clustering algorithm
--> 751 self.fit(X)
752 return self.labels_
753

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/skfda/ml/clustering/_kmeans.py in fit(self, X, y, sample_weight)
327 n_iter,
328 ) = (
--> 329 self._algorithm(
330 fdata=fdata,
331 random_state=random_state,

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/skfda/ml/clustering/_kmeans.py in _algorithm(self, fdata, random_state)
266 centroids_old.data_matrix[...] = centroids.data_matrix
267
--> 268 distances_to_centroids = pairwise_metric(fdata, centroids)
269
270 self._update(

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/skfda/misc/metrics/_utils.py in call(self, elem1, elem2)
181 ) -> NDArrayFloat:
182 """Evaluate the pairwise metric."""
--> 183 optimized = pairwise_metric_optimization(self.metric, elem1, elem2)
184
185 return (

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/multimethod/init.py in call(self, *args, **kwargs)
419 """Resolve and dispatch to best method."""
420 params = self.signature.bind(*args, **kwargs).args if (kwargs and self.signature) else args
--> 421 func = self.dispatch(*params)
422 return func(*args, **kwargs)
423

~/miniconda3/envs/pytorch/lib/python3.9/site-packages/multimethod/init.py in dispatch(self, *args)
352 def dispatch(self, *args) -> Callable:
353 types = tuple(map(type, args))
--> 354 if not any(map(issubclass, types, self.generics)):
355 return self[types]
356 matches = {key for key in list(self) if isinstance(key, signature) and key.instances(*args)}

~/miniconda3/envs/pytorch/lib/python3.9/typing.py in subclasscheck(self, cls)
716
717 def subclasscheck(self, cls):
--> 718 raise TypeError("Subscripted generics cannot be used with"
719 " class and instance checks")
720

TypeError: Subscripted generics cannot be used with class and instance checks

Code to reproduce the bug

In your official example, https://fda.readthedocs.io/en/stable/auto_examples/full_examples/plot_aemet_unsupervised.html#sphx-glr-auto-examples-full-examples-plot-aemet-unsupervised-py


n_clusters = 5
n_init = 10

fda_kmeans = KMeans(
    n_clusters=n_clusters,
    n_init=n_init,
    metric=l2_distance,
    random_state=0,
)
fda_clusters = fda_kmeans.fit_predict(X)

Expected result

successfully work

Actual result

optimization error reported

Traceback (if an exception is raised)

No response

Software versions

scikit-fda version:0.9
OS: MAC 13.2.1 (22D68)

Additional context

No response

[vnmabus]

This is due to an error in one of our dependencies (coady/multimethod#113). While this is being investigated, I recommend installing an older version of multimethod:

pip install multimethod==1.10