Issue 954: Add various data sources from Child Opportunity Index (#986)

* Adds four fields:
    * Summer days above 90F
    * Percent low access to healthy food
    * Percent impenetrable surface areas
    * Low third grade reading proficiency

* Each of these four gets added into Definition L in various factors.

* Additionally, I add college attendance fields to the ETL for Census ACS.

* This PR also introduces the notion of "reverse percentiles", relevant to ticket #970.

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

@@ -1,4 +1,6 @@
 import functools
+from collections import namedtuple
+
 import pandas as pd
 
 from data_pipeline.etl.base import ExtractTransformLoad
@@ -29,6 +31,7 @@ class ScoreETL(ExtractTransformLoad):
         self.persistent_poverty_df: pd.DataFrame
         self.census_decennial_df: pd.DataFrame
         self.census_2010_df: pd.DataFrame
+        self.child_opportunity_index_df: pd.DataFrame
 
     def extract(self) -> None:
         logger.info("Loading data sets from disk.")
@@ -162,6 +165,19 @@ class ScoreETL(ExtractTransformLoad):
             low_memory=False,
         )
 
+        # Load COI data
+        child_opportunity_index_csv = (
+            constants.DATA_PATH
+            / "dataset"
+            / "child_opportunity_index"
+            / "usa.csv"
+        )
+        self.child_opportunity_index_df = pd.read_csv(
+            child_opportunity_index_csv,
+            dtype={self.GEOID_TRACT_FIELD_NAME: "string"},
+            low_memory=False,
+        )
+
     def _join_tract_dfs(self, census_tract_dfs: list) -> pd.DataFrame:
         logger.info("Joining Census Tract dataframes")
 
@@ -255,6 +271,7 @@ class ScoreETL(ExtractTransformLoad):
             self.census_acs_median_incomes_df,
             self.census_decennial_df,
             self.census_2010_df,
+            self.child_opportunity_index_df,
         ]
 
         # Sanity check each data frame before merging.
@@ -323,6 +340,7 @@ class ScoreETL(ExtractTransformLoad):
             field_names.HIGH_SCHOOL_ED_FIELD,
             field_names.UNEMPLOYMENT_FIELD,
             field_names.MEDIAN_HOUSE_VALUE_FIELD,
+            field_names.COLLEGE_ATTENDANCE_FIELD,
             field_names.EXPECTED_BUILDING_LOSS_RATE_FIELD,
             field_names.EXPECTED_AGRICULTURE_LOSS_RATE_FIELD,
             field_names.EXPECTED_POPULATION_LOSS_RATE_FIELD,
@@ -333,6 +351,9 @@ class ScoreETL(ExtractTransformLoad):
             field_names.CENSUS_POVERTY_LESS_THAN_100_FPL_FIELD_2010,
             field_names.CENSUS_DECENNIAL_TOTAL_POPULATION_FIELD_2009,
             field_names.CENSUS_DECENNIAL_AREA_MEDIAN_INCOME_PERCENT_FIELD_2009,
+            field_names.EXTREME_HEAT_FIELD,
+            field_names.HEALTHY_FOOD_FIELD,
+            field_names.IMPENETRABLE_SURFACES_FIELD,
         ]
 
         non_numeric_columns = [
@@ -340,7 +361,32 @@ class ScoreETL(ExtractTransformLoad):
             field_names.PERSISTENT_POVERTY_FIELD,
         ]
 
-        columns_to_keep = non_numeric_columns + numeric_columns
+        # For some columns, high values are "good", so we want to reverse the percentile
+        # so that high values are "bad" and any scoring logic can still check if it's
+        # >= some threshold.
+        # TODO: Add more fields here.
+        #  https://github.com/usds/justice40-tool/issues/970
+        ReversePercentile = namedtuple(
+            typename="ReversePercentile",
+            field_names=["field_name", "low_field_name"],
+        )
+        reverse_percentiles = [
+            # This dictionary follows the format:
+            # <field name> : <field name for low values>
+            # for instance, 3rd grade reading level : Low 3rd grade reading level.
+            # This low field will not exist yet, it is only calculated for the
+            # percentile.
+            ReversePercentile(
+                field_name=field_names.READING_FIELD,
+                low_field_name=field_names.LOW_READING_FIELD,
+            )
+        ]
+
+        columns_to_keep = (
+            non_numeric_columns
+            + numeric_columns
+            + [rp.field_name for rp in reverse_percentiles]
+        )
 
         df_copy = df[columns_to_keep].copy()
 
@@ -375,6 +421,19 @@ class ScoreETL(ExtractTransformLoad):
                 df_copy[col] - min_value
             ) / (max_value - min_value)
 
+        # Create reversed percentiles for these fields
+        for reverse_percentile in reverse_percentiles:
+            # Calculate reverse percentiles
+            # For instance, for 3rd grade reading level (score from 0-500),
+            # calculate reversed percentiles and give the result the name
+            # `Low 3rd grade reading level (percentile)`.
+            df_copy[
+                f"{reverse_percentile.low_field_name}"
+                f"{field_names.PERCENTILE_FIELD_SUFFIX}"
+            ] = df_copy[reverse_percentile.field_name].rank(
+                pct=True, ascending=False
+            )
+
         # Special logic: create a combined population field.
         # We sometimes run analytics on "population", and this makes a single field
         # that is either the island area's population in 2009 or the state's