Updating higher education to be reversed (#1387)

Summary In this PR, we create a new variable so that the % college students is expressed as % not college students. This means that the front end can display % not college students.

Includes old variables so that this will not break fe.

data/data-pipeline/data_pipeline/content/config/csv.yml

@@ -248,3 +248,6 @@ fields:
   - score_name: Greater than or equal to the 90th percentile for low median household income as a percent of area median income and has low HS education in 2009 (island areas)?
     label: Greater than or equal to the 90th percentile for low median household income as a percent of area median income and has low HS education in 2009 (island areas)?
     format: bool
+  - score_name: Percent of population not currently enrolled in college or graduate school
+    label: Percent of residents who are not currently enrolled in higher ed
+    format: percentage

data/data-pipeline/data_pipeline/content/config/excel.yml

@@ -238,6 +238,9 @@ sheets:
       - score_name: Percent individuals age 25 or over with less than high school degree
         label: Percent individuals age 25 or over with less than high school degree
         format: percentage
+      - score_name: Percent of population not currently enrolled in college or graduate school
+        label: Percent of residents who are not currently enrolled in higher ed
+        format: percentage
       - score_name: Unemployment (percent) in 2009 (island areas) and 2010 (states and PR)
         label: Unemployment (percent) in 2009 (island areas) and 2010 (states and PR)
         format: percentage

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

@@ -1,6 +1,5 @@
 from pathlib import Path
 import datetime
-
 from data_pipeline.config import settings
 
 from data_pipeline.score import field_names
@@ -205,6 +204,8 @@ TILES_SCORE_COLUMNS = {
     # Percentage of HS Degree completion for Islands
     field_names.CENSUS_DECENNIAL_HIGH_SCHOOL_ED_FIELD_2009: "IAHSEF",
     field_names.COLLEGE_ATTENDANCE_FIELD: "CA",
+    field_names.COLLEGE_NON_ATTENDANCE_FIELD: "NCA",
+    # This is logically equivalent to "non-college greater than 80%"
     field_names.COLLEGE_ATTENDANCE_LESS_THAN_20_FIELD: "CA_LT20",
     field_names.LOW_INCOME_THRESHOLD: "FPL200S",
     # Booleans for the front end about the types of thresholds exceeded
@@ -270,5 +271,6 @@ TILES_SCORE_FLOAT_COLUMNS = [
     field_names.ISLAND_AREAS_LOW_HS_EDUCATION_FIELD,
     field_names.WASTEWATER_FIELD + field_names.PERCENTILE_FIELD_SUFFIX,
     field_names.SCORE_M + field_names.PERCENTILE_FIELD_SUFFIX,
+    field_names.COLLEGE_NON_ATTENDANCE_FIELD,
     field_names.COLLEGE_ATTENDANCE_FIELD,
 ]

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

@@ -443,6 +443,7 @@ class ScoreETL(ExtractTransformLoad):
             field_names.UNEMPLOYMENT_FIELD,
             field_names.MEDIAN_HOUSE_VALUE_FIELD,
             field_names.COLLEGE_ATTENDANCE_FIELD,
+            field_names.COLLEGE_NON_ATTENDANCE_FIELD,
             field_names.EXPECTED_BUILDING_LOSS_RATE_FIELD,
             field_names.EXPECTED_AGRICULTURE_LOSS_RATE_FIELD,
             field_names.EXPECTED_POPULATION_LOSS_RATE_FIELD,

data/data-pipeline/data_pipeline/etl/sources/census_acs/etl.py

@@ -136,6 +136,8 @@ class CensusACSETL(ExtractTransformLoad):
             "Percent enrollment in college or graduate school"
         )
 
+        self.COLLEGE_NON_ATTENDANCE_FIELD = "Percent of population not currently enrolled in college or graduate school"
+
         self.RE_FIELDS = [
             "B02001_001E",
             "B02001_002E",
@@ -190,6 +192,7 @@ class CensusACSETL(ExtractTransformLoad):
                 self.MEDIAN_HOUSE_VALUE_FIELD_NAME,
                 self.HIGH_SCHOOL_ED_FIELD,
                 self.COLLEGE_ATTENDANCE_FIELD,
+                self.COLLEGE_NON_ATTENDANCE_FIELD,
             ]
             + self.RE_OUTPUT_FIELDS
             + [self.PERCENT_PREFIX + field for field in self.RE_OUTPUT_FIELDS]
@@ -354,6 +357,10 @@ class CensusACSETL(ExtractTransformLoad):
             + df[self.COLLEGE_ATTENDANCE_FEMALE_ENROLLED_PRIVATE]
         ) / df[self.COLLEGE_ATTENDANCE_TOTAL_POPULATION_ASKED]
 
+        df[self.COLLEGE_NON_ATTENDANCE_FIELD] = (
+            1 - df[self.COLLEGE_ATTENDANCE_FIELD]
+        )
+
         # strip columns
         df = df[self.COLUMNS_TO_KEEP]
 

data/data-pipeline/data_pipeline/score/field_names.py

@@ -112,6 +112,9 @@ MEDIAN_INCOME_AS_PERCENT_OF_STATE_FIELD = (
 PERSISTENT_POVERTY_FIELD = "Persistent Poverty Census Tract"
 AMI_FIELD = "Area Median Income (State or metropolitan)"
 COLLEGE_ATTENDANCE_FIELD = "Percent enrollment in college or graduate school"
+COLLEGE_NON_ATTENDANCE_FIELD = (
+    "Percent of population not currently enrolled in college or graduate school"
+)
 MEDIAN_INCOME_AS_PERCENT_OF_AMI_FIELD = (
     "Median household income as a percent of area median income"
 )

data/data-pipeline/data_pipeline/score/score_m.py

@@ -797,9 +797,15 @@ class ScoreM(Score):
             >= self.LOW_INCOME_THRESHOLD
         )
 
+        # Because we are moving this variable to be in the same direction as all
+        # other variables, we change this to be < rather than <=. This translates
+        # to "80% or more of residents are not college students", rather than
+        # "Strictly greater than 80% of residents are not college students."
+        # There are two tracts that are impacted by this (that is, they have exactly)
+        # 20% college students -- neither of these has been a DAC under any score.
         self.df[field_names.COLLEGE_ATTENDANCE_LESS_THAN_20_FIELD] = (
             self.df[field_names.COLLEGE_ATTENDANCE_FIELD]
-            <= self.MAX_COLLEGE_ATTENDANCE_THRESHOLD
+            < self.MAX_COLLEGE_ATTENDANCE_THRESHOLD
         )
 
         self.df[