print Docker image size and fix CSV test

Author: colourmeamused

.github/workflows/test.yml

@@ -22,6 +22,7 @@ jobs:
     - name: Build dea-conflux image
       run: |
         docker compose build
+        docker image inspect "$(docker compose images -q conflux)" --format='dea-conflux image size: {{.Size}} bytes'
 
     - name: Setup dea-conflux test environment
       run: |

dea_conflux/stack.py

@@ -281,7 +281,7 @@ def save_df_as_csv(single_polygon_df, feature_id, outpath, remove_duplicated_dat
         # Remove the timeseries duplicated data
         single_polygon_df = remove_timeseries_with_duplicated(single_polygon_df)
     single_polygon_df["nwi_id"] = single_polygon_df.index
-    single_polygon_df.reset_index(inplace=True)
+    single_polygon_df.reset_index(drop=True, inplace=True)
 
     # WIT Normalise Step
 
@@ -319,9 +319,8 @@ def save_df_as_csv(single_polygon_df, feature_id, outpath, remove_duplicated_dat
     single_polygon_df = dea_tools.wetlands.generate_low_quality_data_periods(single_polygon_df)
     print(single_polygon_df)
     dea_tools.wetlands.display_wit_stack_with_df(single_polygon_df, feature_id, feature_id, x_axis_labels="years")
-    # Drop only the temporary reset_index column; veg_areas and overall_veg_num
-    # are retained as they are defined in the WIT data dictionary.
-    single_polygon_df.drop(["index"], axis=1, inplace=True)
+    # veg_areas and overall_veg_num are retained as they are defined in the WIT
+    # data dictionary.
     single_polygon_df.rename(columns=WIT_CSV_COLUMN_RENAMES, inplace=True)
     # Reorder to match WIT_CSV_COLUMN_ORDER; any unlisted columns go at the end.
     ordered = [c for c in WIT_CSV_COLUMN_ORDER if c in single_polygon_df.columns]

tests/test_stack.py

@@ -119,14 +119,23 @@ def test_wit_csv_column_names_match_data_dictionary(tmp_path, monkeypatch):
     )
 
     csv = pd.read_csv(output)
-    assert {
+    assert list(csv.columns) == [
         "nwi_id",
+        "date",
         "green_vegetation",
         "dry_vegetation",
         "bare_soil",
         "wetness",
-    }.issubset(csv.columns)
-    assert not {"feature_id", "pv", "npv", "bs", "wet"}.intersection(csv.columns)
+        "water",
+        "veg_areas",
+        "overall_veg_num",
+        "norm_bs",
+        "norm_pv",
+        "norm_npv",
+        "off_value",
+        "pc_missing",
+    ]
+    assert csv.loc[0, "nwi_id"] == WIT_POLYGON_ID
 
 
 # def test_wit_stacking(tmp_path):