fixed column ordering

Author: wjm41

molplotly/main.py

@@ -50,6 +50,15 @@ def test_groups(fig: Figure, df_grouped: DataFrameGroupBy):
     return True
 
 
+def find_correct_column_order(cols, col_names):
+
+    correctly_ordered_cols = []
+    for col in col_names:
+        if col in cols:
+            correctly_ordered_cols.append(col)
+    return correctly_ordered_cols
+
+
 def find_grouping(
     fig: Figure, df_data: pd.DataFrame, cols: list[str]
 ) -> tuple[DataFrameGroupBy, dict]:
@@ -62,6 +71,7 @@ def find_grouping(
                 f"marker_col/color_col/facet_col is misspecified because the specified dataframe grouping names {cols} don't match the names in the plotly figure {col_names}.",
             )
 
+        cols = find_correct_column_order(cols, col_names)
         df_grouped = df_data.groupby(cols)
 
         str_groups = {}
@@ -77,16 +87,24 @@ def find_grouping(
             curve_name = ", ".join(str(x) for x in curve_name)
             if "%{x}" in curve_name:
                 unique_x_values = np.unique(data.x)
-                if len(unique_x_values) == 1:
+                if len(unique_x_values) == 1 and len(data.x) != 1:
                     curve_name = curve_name.replace("%{x}", str(unique_x_values[0]))
                 else:
                     curve_name = curve_name.replace(", %{x}", "")
             if "%{y}" in curve_name:
                 unique_y_values = np.unique(data.y)
-                if len(unique_y_values) == 1:
+                if len(unique_y_values) == 1 and len(data.y) != 1:
                     curve_name = curve_name.replace("%{y}", str(unique_y_values[0]))
                 else:
                     curve_name = curve_name.replace(", %{y}", "")
+            if "%{marker.size}" in curve_name:
+                unique_size_values = np.unique(data.marker.size)
+                if len(unique_size_values) == 1 and len(data.marker.size) != 1:
+                    curve_name = curve_name.replace(
+                        "%{marker.size}", str(unique_size_values[0])
+                    )
+                else:
+                    curve_name = curve_name.replace(", %{marker.size}", "")
             curve_dict[index] = str_groups[curve_name]
 
         return df_grouped, curve_dict
@@ -182,8 +200,6 @@ def add_molecules(
         name of the column in df that is used to color the datapoints in df - necessary when there is discrete conditional coloring (default None).
     symbol_col : str, optional
         name of the column in df that is used to determine the symbols of the datapoints in df (default None).
-    marker_col : str, optional
-        name of the column in df that is used to determine the marker shape of the datapoints in df (default None).
     facet_col : str, optional
         name of the column in df that is used to facet the data to multiple plots (default None).
     wrap : bool, optional
@@ -202,8 +218,6 @@ def add_molecules(
         df_data[color_col] = df_data[color_col].astype(str)
     if symbol_col is not None:
         df_data[symbol_col] = df_data[symbol_col].astype(str)
-    if marker_col is not None:
-        df_data[marker_col] = df_data[marker_col].astype(str)
     if facet_col is not None:
         df_data[facet_col] = df_data[facet_col].astype(str)
 
@@ -214,8 +228,6 @@ def add_molecules(
 
         if color_col is not None:
             cols.append(color_col)
-        if marker_col is not None:
-            cols.append(marker_col)
         if symbol_col is not None:
             cols.append(symbol_col)
         if facet_col is not None: