I can initialize a hotspot object with hs = hotspot.Hotspot(counts=counts,model='danb', latent=pca_res) even if type(pca_res) is numpy.ndarray. No error is thrown, but then hs.create_knn_graph(weighted_graph=False, n_neighbors=30) throws a fairly uninformative error. The code works fine as long as I cast everything to pandas DataFrame before initializing. Using version 0.9.0.
I can initialize a hotspot object with
hs = hotspot.Hotspot(counts=counts,model='danb', latent=pca_res)even iftype(pca_res)isnumpy.ndarray. No error is thrown, but thenhs.create_knn_graph(weighted_graph=False, n_neighbors=30)throws a fairly uninformative error. The code works fine as long as I cast everything topandas DataFramebefore initializing. Using version0.9.0.