Resolving #532

www/api/resources/plotly_data.py

@@ -314,11 +314,12 @@ def post(self, dataset_id):
         if color_map and color_name:
             # Validate if all color map keys are in the dataframe columns
             # Ran into an issue where the color map keys were truncated compared to the dataframe column values
-            col_values = set(df[color_name].unique())
+            # Setting to categories ensures nulls are dropped, which are dropped already in the dataframe
+            col_values = set(df[color_name].cat.categories)
             diff = col_values.difference(color_map.keys())
             if diff:
                 message =  "WARNING: Color map has values not in the dataframe column '{}': {}\n".format(color_name, diff)
-                message += "Will set color map key values to the unique values in the dataframe column."
+                message += "Will set color map key values to the unique Categorical values in the dataframe column."
                 print(message, file=sys.stderr)
                 # Sort both the colormap and dataframe column alphabetically
                 sorted_column_values = sorted(col_values)
@@ -345,7 +346,7 @@ def post(self, dataset_id):
                 ]
                 color_map = purples
             else:
-                names = df[color_name].unique().tolist()
+                names = df[color_name].cat.categories.tolist()
                 color_map = plotly_color_map(names)
 
                 # Check if color hexcodes exist and use if validated
@@ -355,7 +356,7 @@ def post(self, dataset_id):
                     # Ensure one-to-one mapping of color names to codes
                     if len(grouped) == len(names):
                         # Test if names are color hexcodes and use those if applicable
-                        color_hex = df[color_code].unique().tolist()
+                        color_hex = df[color_code].cat.categories.tolist()
                         if re.search(COLOR_HEX_PTRN, color_hex[0]):
                             color_map = {name[0]:name[1] for name, group in grouped}