cmap=cmap, node_color= ) nx.draw_networkx_edges(G, pos, alpha=0.5) plt.show() #community detection communityListWeeks = [] for w in range(0,10): print('Week ' + str(w) + '...') num_comms = len(graph_all_weeks[w].graph._node) communityListWeeks.append(graphLearning.community_dection_graph(graph_all_weeks[w], num_comms=num_comms)) import graphLearning import scikit_posthocs as sp pd.set_option("display.max_rows", None, "display.max_columns", None) aw10 = graphLearning.extractAssessmentResultOfCommunities(communityListWeeks[9], assessment3A, 'perCorrect3A') aw10t = sp.posthoc_conover(aw10[3][5]) goodCommunity = aw10[3][5][0] badCommunity = aw10[3][5][1] w = 9 for w in range(0,10): activityDataMatrixWeeks_pageTypeWeek[w] = graphLearning.mapNewLabel(activityDataMatrixWeeks_pageTypeWeek[w] , reLabelIndex) extractGoodBadCommunity = activityDataMatrixWeeks_pageTypeWeek[w].loc[activityDataMatrixWeeks_pageTypeWeek[w].index.astype(int).isin(goodCommunity.index) | activityDataMatrixWeeks_pageTypeWeek[w].index.astype(int).isin(badCommunity.index)] extractGoodBadCommunity['group'] = 0 extractGoodBadCommunity.loc[extractGoodBadCommunity.index.astype(int).isin(goodCommunity.index),['group']] = 0 extractGoodBadCommunity.loc[extractGoodBadCommunity.index.astype(int).isin(badCommunity.index),['group']] = 1 columnListStatsSig = [] for c in extractGoodBadCommunity.columns:
num_comms=num_comms)) communityListWeeks_cleaned_correlation = [] for w in range(0, 12): print('Week ' + str(w) + '...') num_comms = len(graph_all_weeks_cleaned_correlation[w].graph._node) communityListWeeks_cleaned_correlation.append( graphLearning.community_dection_graph( graph_all_weeks_cleaned_correlation[w], num_comms=num_comms)) import graphLearning import scikit_posthocs as sp pd.set_option("display.max_rows", None, "display.max_columns", None) aw10 = graphLearning.extractAssessmentResultOfCommunities( communityListWeeks_cleaned_correlation[11], assessment2A, 'perCorrect2A') aw10t = sp.posthoc_conover(aw10[1][5]) aw6 = graphLearning.extractAssessmentResultOfCommunities( communityListWeeks[6], assessment1A, 'perCorrect1A') aw6t = sp.posthoc_conover(aw6[3][5]) a = graphLearning.findTogetherMembers(aw10[18][5], aw11[18][5], aw10[18][1], aw11[18][1]) len(set(assessment2A.index).intersection(set(assessment3A.index))) a = graphLearning.findTogetherMembers(aw7[18][5], aw11[18][5], aw7[18][1], aw11[18][1]) for i in range(0, 8): for j in range(0, 8): if len(a[i][j]) > 1: print(a[i][j])
print('Week ' + str(w) + '...') num_comms = len(graph_all_weeks[w].graph._node) communityListWeeks.append(graphLearning.community_dection_graph(graph_all_weeks[w], most_valuable_edge=graphLearning.most_central_edge, num_comms=num_comms)) communityListWeeks_not_cleaned = [] for w in range(0,12): print('Week ' + str(w) + '...') num_comms = len(graph_all_weeks_not_cleaned[w].graph._node) communityListWeeks_not_cleaned.append(graphLearning.community_dection_graph(graph_all_weeks_not_cleaned[w], most_valuable_edge=graphLearning.most_central_edge, num_comms=num_comms)) import graphLearning import scikit_posthocs as sp pd.set_option("display.max_rows", None, "display.max_columns", None) aw10 = graphLearning.extractAssessmentResultOfCommunities(communityListWeeks[10], assessment2A, 'perCorrect2A') aw10t = sp.posthoc_conover(aw10[0][5]) aw6 = graphLearning.extractAssessmentResultOfCommunities(communityListWeeks[6], assessment1A, 'perCorrect1A') aw6t = sp.posthoc_conover(aw7[3][5]) a = graphLearning.findTogetherMembers(aw10[18][5],aw11[18][5], aw10[18][1],aw11[18][1]) len(set(assessment2A.index).intersection(set(assessment3A.index))) a = graphLearning.findTogetherMembers(aw7[18][5],aw11[18][5], aw7[18][1],aw11[18][1]) for i in range(0,8): for j in range(0,8): if len(a[i][j]) > 1: print(a[i][j])