BUG: preserve column order inside iterfeatures (#2396)

geopandas/geodataframe.py

@@ -865,7 +865,7 @@ individually so that features may have different properties
         if not self.columns.is_unique:
             raise ValueError("GeoDataFrame cannot contain duplicated column names.")
 
-        properties_cols = self.columns.difference([self._geometry_column_name])
+        properties_cols = self.columns.drop(self._geometry_column_name)
 
         if len(properties_cols) > 0:
             # convert to object to get python scalars.

geopandas/tests/test_geodataframe.py

@@ -659,7 +659,7 @@ class TestDataFrame:
         geometry = [Point(xy) for xy in zip(df["lon"], df["lat"])]
         gdf = GeoDataFrame(df, geometry=geometry)
         # from_features returns sorted columns
-        expected = gdf[["geometry", "lat", "lon", "name"]]
+        expected = gdf[["geometry", "name", "lat", "lon"]]
 
         # test FeatureCollection
         res = GeoDataFrame.from_features(gdf.__geo_interface__)