T[i_upper] distance_matrix = np.subtract(1,similarity_matrix) TSNE_sim = TSNE(verbose=1, n_components=2, init='pca', method='barnes_hut', perplexity=perp).fit_transform(distance_matrix) tsne_result = pd.DataFrame(data = TSNE_sim , columns=["TC1","TC2"]) return tsne_result sub_...
parameter can be used to construct a heterogeneous graph from a single flat DataFrame, containing a column of the edge types #1284. This avoids the need to build separate DataFrames for each type, and is significantly faster when there are many types. Using edge_type_column gives a 2.6× ...