I'm trying to create a heatmap to show relationships between how many times Customers use reports, using the Count column as the values within the Heatmap Customers can use several templates as many times as they want. My dataframe looks something like this (Original data frame has 638 rows and I have created dummy data to conceal the original data for privacy purposes): print(df) Customer Report Count 1 XIAA Report_A 198 2 XIAA Report_C 1 3 BKAN Bespoke 65 4 JAJS Report_A 3 5 HNUA Report_F 436 6 NOAO Bespoke 12 7 RQNJ Report_G 53 8 MJAS Report_A 9 9 PGBC Report_Y 155 To plot a heatmap, I'm using the function: heatmap1_data = pd.pivot_table(df, values='Count', index=['Customer'], columns='Report') plt.subplots(figsize=(25, 30)) sns.heatmap(heatmap1_data, cmap="Blues",vmin=0, vmax=2) My heatmap is returned like this (I have cropped the names in the axis for privacy purposes): The heatmap looks too separated due to the variation of the counts in my dataframe, which causes the Pivot table to have many '0' values to show which Reports are not being used by certain customers. Is there a way to sort this so that the highest correlations are shown together within the heatmap instead of them being very seperated? Continue reading...
