Reformat NSSP county zip code (#1976)

* Format county fips to all be 5 digits with leading zeros + Add test

* Update test_pull.py wordings

* linter

Author: minhkhul

nssp/delphi_nssp/constants.py

@@ -37,6 +37,6 @@
         "timestamp": "datetime64[ns]",
         "geography": str,
         "county": str,
-        "fips": int,
+        "fips": str,
     }
 )

nssp/delphi_nssp/pull.py

@@ -67,5 +67,8 @@ def pull_nssp_data(socrata_token: str):
     except KeyError as exc:
         raise ValueError(warn_string(df_ervisits, TYPE_DICT)) from exc
 
+    # Format county fips to all be 5 digits with leading zeros
+    df_ervisits["fips"] = df_ervisits["fips"].apply(lambda x: str(x).zfill(5) if str(x) != "0" else "0")
+
     keep_columns = ["timestamp", "geography", "county", "fips"]
     return df_ervisits[SIGNALS + keep_columns]

nssp/tests/test_data/page.txt

@@ -61,5 +61,26 @@
         "hsa_nci_id": "All",
         "fips": "0",
         "trend_source": "United States"
+    },
+    {
+    "week_end": "2023-05-13T00:00:00.000",
+    "geography": "Colorado",
+    "county": "Jefferson",
+    "percent_visits_combined": "0.84",
+    "percent_visits_covid": "0.59",
+    "percent_visits_influenza": "0.23",
+    "percent_visits_rsv": "0.03",
+    "percent_visits_smoothed": "0.83",
+    "percent_visits_smoothed_covid": "0.62",
+    "percent_visits_smoothed_1": "0.18",
+    "percent_visits_smoothed_rsv": "0.02",
+    "ed_trends_covid": "Decreasing",
+    "ed_trends_influenza": "No Change",
+    "ed_trends_rsv": "Decreasing",
+    "hsa": "Denver (Denver), CO - Jefferson, CO",
+    "hsa_counties": "Adams, Arapahoe, Clear Creek, Denver, Douglas, Elbert, Gilpin, Grand, Jefferson, Park, Summit",
+    "hsa_nci_id": "688",
+    "fips": "8059",
+    "trend_source": "HSA"
     }
 ]

nssp/tests/test_pull.py

@@ -49,6 +49,7 @@ def test_pull_nssp_data(self, mock_socrata):
         assert result["geography"].notnull().all(), "geography has rogue NaN"
         assert result["county"].notnull().all(), "county has rogue NaN"
         assert result["fips"].notnull().all(), "fips has rogue NaN"
+        assert result["fips"].apply(lambda x: isinstance(x, str) and len(x) != 4).all(), "fips formatting should always be 5 digits; include leading zeros if aplicable"
 
         # Check for each signal in SIGNALS
         for signal in SIGNALS: