sktime: [BUG] Conversion from_multi_index_to_3d_numpy(X=obj)

Describe the bug

I’d like to start out by saying this is probably an issue with my code, however, whenever I am running basic models from the examples I keep coming into a reshape error.

for instance I first run my data through check_raise(X_train,mtype=“pd-multiindex”) and it returns true. For this example I made a time series with 38 instances (subjects) with 1 time point each and 2 columns. Note the math is the same with real data.

To Reproduce

I run the code: from sktime.classification.distance_based import KNeighborsTimeSeriesClassifier classifier = KNeighborsTimeSeriesClassifier(distance="euclidean" , n_jobs=-1) pipe = make_pipeline(classifier) pipe.fit(X_train, y_train) y_pred = pipe.predict(X_test) accuracy_score(y_test, y_pred)

and it returns: ValueError: cannot reshape array of size 76 into shape (38,38,2)

I also took another example this time using: clf = ColumnEnsembleClassifier( estimators=[ ("TSF0", TimeSeriesForestClassifier(n_estimators=100), [0]) ] ) pipe = make_pipeline(clf) pipe.fit(X_train, y_train) y_pred = pipe.predict(X_test) accuracy_score(y_test, y_pred)

and it returns: ValueError: cannot reshape array of size 38 into shape (38,38,1)

Expected behavior

I thought it would run as it passed the check.

Additional context

When I add in default padding pad = PaddingTransformer(pad_length=None, fill_value=0) both models works which is strange as I am unsure what they are padding if all of my sequences are truncated to 1 timepoint each

Versions

System: python: 3.10.7 (v3.10.7:6cc6b13308, Sep 5 2022, 14:02:52) [Clang 13.0.0 (clang-1300.0.29.30)] executable: /Library/Frameworks/Python.framework/Versions/3.10/bin/python3.10 machine: macOS-12.0.1-arm64-arm-64bit

Python dependencies: pip: 22.2.2 setuptools: 59.8.0 sklearn: 1.1.2 sktime: 0.13.4 statsmodels: 0.13.2 numpy: 1.22.4 scipy: 1.8.1 pandas: 1.4.4 matplotlib: 3.6.0 joblib: 1.2.0 numba: 0.56.2 pmdarima: None tsfresh: None