Fitting the k-means-constrained on training samples and predicting on test samples raises error

[bweill555]

Hi,
I'm trying to fit the k-means-constrained on training samples and then call it to predict test samples. I am getting the following error message:

~\anaconda3\lib\site-packages\k_means_constrained\k_means_constrained_.py in predict(self, X, size_min, size_max)
708 raise ValueError("size_max must be larger than size_min")
709 if size_min * n_clusters > n_samples:
--> 710 raise ValueError("The product of size_min and n_clusters cannot exceed the number of samples (X)")
711
712 labels, inertia = \

ValueError: The product of size_min and n_clusters cannot exceed the number of samples (X)

It seems there is not enough data in the testing sample to meet the clusters size constraints (here size_min) but is there a way to only apply the clusters sizes constrains in the fitting process and not in the prediction one?

[joshlk]

Hey, thanks for using k-means-constrained! Can you give show me the full code example of what you are doing and the shapes of the input data?

[joshlk]

I’m going to close the issue due to inactivity. @bweill555 feel free to reopen if you are still having issues