Improve score test documentation based on Lucas's feedback (#1835) (#1914)

* Better document base on Lucas's feedback (#1835)

* Fix typo (#1835)

* Add test to verify GEOJSON matches tiles (#1835)

* Remove NOOP line (#1835)

* Move GEOJSON generation up for new smoketest (#1835)

* Fixup code format (#1835)

* Update readme for new somketest (#1835)

.github/workflows/deploy_be_staging.yml

@@ -61,7 +61,10 @@ jobs:
           poetry run python3 data_pipeline/application.py score-full-run
       - name: Generate Score Post
         run: |
-          poetry run python3 data_pipeline/application.py generate-score-post -s aws
+          poetry run python3 data_pipeline/application.py generate-score-post
+      - name: Generate Score Geo
+        run: |
+          poetry run python3 data_pipeline/application.py geo-score
       - name: Run Smoketests
         run: |
           poetry run pytest data_pipeline/ -m smoketest
@@ -100,9 +103,6 @@ jobs:
           mkdir -p /usr/local/bin
           cp tippecanoe /usr/local/bin/tippecanoe
           tippecanoe -v
-      - name: Generate Score Geo
-        run: |
-          poetry run python3 data_pipeline/application.py geo-score
       - name: Generate Tiles
         run: |
           poetry run python3 data_pipeline/application.py generate-map-tiles

data/data-pipeline/README.md

@@ -12,11 +12,14 @@
       - [2. Extract-Transform-Load (ETL) the data](#2-extract-transform-load-etl-the-data)
       - [3. Combined dataset](#3-combined-dataset)
       - [4. Tileset](#4-tileset)
+      - [5. Shapefiles](#5-shapefiles)
     - [Score generation and comparison workflow](#score-generation-and-comparison-workflow)
       - [Workflow Diagram](#workflow-diagram)
       - [Step 0: Set up your environment](#step-0-set-up-your-environment)
       - [Step 1: Run the script to download census data or download from the Justice40 S3 URL](#step-1-run-the-script-to-download-census-data-or-download-from-the-justice40-s3-url)
       - [Step 2: Run the ETL script for each data source](#step-2-run-the-etl-script-for-each-data-source)
+        - [Table of commands](#table-of-commands)
+        - [ETL steps](#etl-steps)
       - [Step 3: Calculate the Justice40 score experiments](#step-3-calculate-the-justice40-score-experiments)
       - [Step 4: Compare the Justice40 score experiments to other indices](#step-4-compare-the-justice40-score-experiments-to-other-indices)
     - [Data Sources](#data-sources)
@@ -26,21 +29,27 @@
     - [MacOS](#macos)
     - [Windows Users](#windows-users)
     - [Setting up Poetry](#setting-up-poetry)
-    - [Downloading Census Block Groups GeoJSON and Generating CBG CSVs](#downloading-census-block-groups-geojson-and-generating-cbg-csvs)
+    - [Running tox](#running-tox)
+    - [The Application entrypoint](#the-application-entrypoint)
+    - [Downloading Census Block Groups GeoJSON and Generating CBG CSVs (not normally required)](#downloading-census-block-groups-geojson-and-generating-cbg-csvs-not-normally-required)
+    - [Run all ETL, score and map generation processes](#run-all-etl-score-and-map-generation-processes)
+    - [Run both ETL and score generation processes](#run-both-etl-and-score-generation-processes)
+    - [Run all ETL processes](#run-all-etl-processes)
     - [Generating Map Tiles](#generating-map-tiles)
     - [Serve the map locally](#serve-the-map-locally)
     - [Running Jupyter notebooks](#running-jupyter-notebooks)
     - [Activating variable-enabled Markdown for Jupyter notebooks](#activating-variable-enabled-markdown-for-jupyter-notebooks)
-  - [Miscellaneous](#miscellaneous)
   - [Testing](#testing)
     - [Background](#background)
-    - [Configuration / Fixtures](#configuration--fixtures)
+    - [Score and post-processing tests](#score-and-post-processing-tests)
       - [Updating Pickles](#updating-pickles)
-      - [Future Enchancements](#future-enchancements)
+      - [Future Enhancements](#future-enhancements)
-    - [ETL Unit Tests](#etl-unit-tests)
+    - [Fixtures used in ETL "snapshot tests"](#fixtures-used-in-etl-snapshot-tests)
+    - [Other ETL Unit Tests](#other-etl-unit-tests)
       - [Extract Tests](#extract-tests)
       - [Transform Tests](#transform-tests)
       - [Load Tests](#load-tests)
+    - [Smoketests](#smoketests)
 
 <!-- /TOC -->
 
@@ -496,3 +505,13 @@ See above [Fixtures](#configuration--fixtures) section for information about whe
 These make use of [tmp_path_factory](https://docs.pytest.org/en/latest/how-to/tmp_path.html) to create a file-system located under `temp_dir`, and validate whether the correct files are written to the correct locations.
 
 Additional future modifications could include the use of Pandera and/or other schema validation tools, and or a more explicit test that the data written to file can be read back in and yield the same dataframe.
+
+### Smoketests
+
+To ensure the score and tiles process correctly, there is a suite of "smoke tests" that can be run after the ETL and score data have been run, and outputs like the frontend GEOJSON have been created.
+These tests are implemented as pytest test, but are skipped by default. To run them.
+
+1. Generate a full score with `poetry run python3 data_pipeline/application.py score-full-run`
+2. Generate the tile data with `poetry run python3 data_pipeline/application.py generate-score-post`
+3. Generate the frontend GEOJSON with `poetry run python3 data_pipeline/application.py geo-score`
+4. Select the smoke tests for pytest with `poetry run pytest data_pipeline/tests -k smoketest`

data/data-pipeline/data_pipeline/etl/score/etl_score_geo.py

@@ -41,7 +41,6 @@ class GeoScoreETL(ExtractTransformLoad):
         self.SCORE_CSV_PATH = self.DATA_PATH / "score" / "csv"
         self.TILE_SCORE_CSV = self.SCORE_CSV_PATH / "tiles" / "usa.csv"
 
-        self.DATA_SOURCE = data_source
         self.CENSUS_USA_GEOJSON = (
             self.DATA_PATH / "census" / "geojson" / "us.json"
         )

data/data-pipeline/data_pipeline/tests/score/test_output.py

@@ -31,7 +31,7 @@ from .fixtures import (
 
 
 pytestmark = pytest.mark.smoketest
-UNMATCHED_TRACK_THRESHOLD = 1000
+UNMATCHED_TRACT_THRESHOLD = 1000
 
 
 def _helper_test_count_exceeding_threshold(df, col, error_check=1000):
@@ -254,6 +254,15 @@ def test_data_sources(
         key: value for key, value in locals().items() if key != "final_score_df"
     }
 
+    # For each data source that's injected via the fixtures, do the following:
+    # * Ensure at least one column from the source shows up in the score
+    # * Ensure any tracts NOT in the data source are NA/null in the score
+    # * Ensure the data source doesn't have a large number of tract IDs that are not
+    #   included in the final score, since that implies the source is using 2020
+    #   tract IDs
+    # * Verify that the data from the source that's in the final score output
+    #   is the "equal" to the data from the ETL, allowing for the minor
+    #   differences that come from floating point comparisons
     for data_source_name, data_source in data_sources.items():
         final = "final_"
         df: pd.DataFrame = final_score_df.merge(
@@ -275,12 +284,12 @@ def test_data_sources(
         ), f"No columns from data source show up in final score in source {data_source_name}"
 
         # Make sure we have NAs for any tracts in the final data that aren't
-        # covered in the  final data
+        # included in the data source
         assert np.all(df[df.MERGE == "left_only"][final_columns].isna())
 
         # Make sure the datasource doesn't have a ton of unmatched tracts, implying it
         # has moved to 2020 tracts
-        assert len(df[df.MERGE == "right_only"]) < UNMATCHED_TRACK_THRESHOLD
+        assert len(df[df.MERGE == "right_only"]) < UNMATCHED_TRACT_THRESHOLD
 
         df = df[df.MERGE == "both"]
 
@@ -293,6 +302,7 @@ def test_data_sources(
                 f"Column {final_column} not equal "
                 f"between {data_source_name} and final score"
             )
+            # For non-numeric types, we can use the built-in equals from pandas
             if df[final_column].dtype in [
                 np.dtype(object),
                 np.dtype(bool),
@@ -301,6 +311,8 @@ def test_data_sources(
                 assert df[final_column].equals(
                     df[data_source_column]
                 ), error_message
+            # For numeric sources, use np.close so we don't get harmed by
+            # float equaity weirdness
             else:
                 assert np.allclose(
                     df[final_column],

data/data-pipeline/data_pipeline/tests/score/test_tiles_smoketests.py

@@ -2,6 +2,7 @@
 from dataclasses import dataclass
 from typing import Optional
 import pandas as pd
+import geopandas as gpd
 import numpy as np
 import pytest
 from data_pipeline.config import settings
@@ -26,6 +27,13 @@ def tiles_df(scope="session"):
     )
 
 
+@pytest.fixture()
+def tiles_geojson_df():
+    return gpd.read_file(
+        settings.APP_ROOT / "data" / "score" / "geojson" / "usa-high.json"
+    )
+
+
 PERCENTILE_FIELDS = [
     "DF_PFS",
     "AF_PFS",
@@ -102,6 +110,19 @@ def test_tract_equality(tiles_df, final_score_df):
     assert tiles_df.shape[0] == final_score_df.shape[0]
 
 
+def is_col_fake_bool(col) -> bool:
+    if col.dtype == np.dtype("float64"):
+        fake_bool = {1.0, 0.0, None}
+        # Replace the nans in the column values with None for
+        # so we can just use issubset below
+        col_values = set(
+            not np.isnan(val) and val or None
+            for val in col.value_counts(dropna=False).index
+        )
+        return len(col_values) <= 3 and col_values.issubset(fake_bool)
+    return False
+
+
 @dataclass
 class ColumnValueComparison:
     final_score_column: pd.Series
@@ -110,16 +131,7 @@ class ColumnValueComparison:
 
     @property
     def _is_tiles_column_fake_bool(self) -> bool:
-        if self.tiles_column.dtype == np.dtype("float64"):
+        return is_col_fake_bool(self.tiles_column)
-            fake_bool = {1.0, 0.0, None}
-            # Replace the nans in the column values with None for
-            # so we can just use issubset below
-            col_values = set(
-                not np.isnan(val) and val or None
-                for val in self.tiles_column.value_counts(dropna=False).index
-            )
-            return len(col_values) <= 3 and col_values.issubset(fake_bool)
-        return False
 
     @property
     def _is_dtype_ok(self) -> bool:
@@ -215,6 +227,49 @@ def test_for_column_fidelitiy_from_score(tiles_df, final_score_df):
     assert not errors, error_message
 
 
+def test_for_geojson_fidelity_from_tiles_csv(tiles_df, tiles_geojson_df):
+    tiles_geojson_df = tiles_geojson_df.drop(columns=["geometry"]).rename(
+        columns={"GEOID10": "GTF"}
+    )
+    assert tiles_df.shape == tiles_geojson_df.shape
+    assert tiles_df["GTF"].equals(tiles_geojson_df["GTF"])
+    assert sorted(tiles_df.columns) == sorted(tiles_geojson_df.columns)
+
+    # Are all the dtypes and values the same?
+    for col_name in tiles_geojson_df.columns:
+        if is_col_fake_bool(tiles_df[col_name]):
+            tiles_df[col_name] = (
+                tiles_df[col_name]
+                .astype("float64")
+                .replace({0.0: False, 1.0: True})
+            )
+        if is_col_fake_bool(tiles_geojson_df[col_name]):
+            tiles_geojson_df[col_name] = (
+                tiles_geojson_df[col_name]
+                .astype("float64")
+                .replace({0.0: False, 1.0: True})
+            )
+        tiles_geojson_df[col_name] = tiles_df[col_name].replace({None: np.nan})
+        error_message = f"Column {col_name} not equal "
+        # For non-numeric types, we can use the built-in equals from pandas
+        if tiles_df[col_name].dtype in [
+            np.dtype(object),
+            np.dtype(bool),
+            np.dtype(str),
+        ]:
+            assert tiles_df[col_name].equals(
+                tiles_geojson_df[col_name]
+            ), error_message
+        # For numeric sources, use np.close so we don't get harmed by
+        # float equaity weirdness
+        else:
+            assert np.allclose(
+                tiles_df[col_name],
+                tiles_geojson_df[col_name],
+                equal_nan=True,
+            ), error_message
+
+
 def test_for_state_names(tiles_df):
     states = tiles_df["SF"].value_counts(dropna=False).index
     assert np.nan not in states