Apply feedback from revieiw (#1780)

usds · Aug 16, 2022 · 702804e · 702804e
1 parent 2aff134
commit 702804e
Show file tree

Hide file tree

Showing 4 changed files with 431 additions and 18 deletions.
diff --git a/data/data-pipeline/data_pipeline/etl/score/config/datasets.yml b/data/data-pipeline/data_pipeline/etl/score/config/datasets.yml
@@ -149,14 +149,14 @@ datasets:
     short_name: "eAMLIS"
     module_name: "eamlis"
     load_fields:
-      - short_name: "has_amlis"
-        df_field_name: "BINARY_HAS_AMLIS"
+      - short_name: "has_aml"
+        df_field_name: "AML_BOOLEAN"
         long_name: "Is there at least one abandoned mine in this census tract?"
         description_short: 
           "Whether the tract has an abandoned mine"
         field_type: bool
-        include_in_tiles: false
-        include_in_downloadable_files: false   
+        include_in_tiles: true
+        include_in_downloadable_files: true
   - long_name: "Example ETL"
     short_name: "Example"
     module_name: "example_dataset"

diff --git a/data/data-pipeline/data_pipeline/etl/sources/eamlis/etl.py b/data/data-pipeline/data_pipeline/etl/sources/eamlis/etl.py
@@ -10,15 +10,15 @@
 logger = get_module_logger(__name__)
 
 
-class AbandonedMineLandInventorySystem(ExtractTransformLoad):
+class AbandonedMineETL(ExtractTransformLoad):
     """Data from Office Of Surface Mining Reclamation and Enforcement's
     eAMLIS. These are the locations of abandoned mines.
     """
 
     # Metadata for the baseclass
     NAME = "eamlis"
     GEO_LEVEL = ValidGeoLevel.CENSUS_TRACT
-    BINARY_HAS_AMLIS: str = "Has abandoned mine"
+    AML_BOOLEAN: str
 
     # Define these for easy code completion
     def __init__(self):
@@ -35,7 +35,7 @@ def __init__(self):
 
         self.COLUMNS_TO_KEEP = [
             self.GEOID_TRACT_FIELD_NAME,
-            self.BINARY_HAS_AMLIS,
+            self.AML_BOOLEAN,
         ]
 
         self.output_df: pd.DataFrame
@@ -55,9 +55,9 @@ def transform(self) -> None:
             ),
             crs="epsg:4326",
         )
-        gdf.drop_duplicates(subset=["geometry"], inplace=True, keep="last")
+        gdf = gdf.drop_duplicates(subset=["geometry"], keep="last")
         gdf_tracts = add_tracts_for_geometries(gdf)
-        gdf_tracts.drop_duplicates(self.GEOID_TRACT_FIELD_NAME, inplace=True)
-        gdf_tracts[self.BINARY_HAS_AMLIS] = True
+        gdf = gdf_tracts.drop_duplicates(self.GEOID_TRACT_FIELD_NAME)
+        gdf_tracts[self.AML_BOOLEAN] = True
         self.output_df = gdf_tracts[self.COLUMNS_TO_KEEP]