def make_crimerate_df(self, crimes):
crimes = pd.DataFrame(crimes,
columns=["date_reported", "agg_area", "count"])
crimes["year_reported"] = crimes["date_reported"].apply(
lambda d: date(d).year)
crimes["month_reported"] = crimes["date_reported"].apply(
lambda d: date(d).month)
crimes = crimes.set_index("agg_area")
return crimes[["year_reported", "month_reported", "count"]]
def make_expected_df(self, expected):
expected = pd.DataFrame(
expected, columns=["parcel_id", "inspection_date", "crime_rate"])
expected["inspection_date"] = expected["inspection_date"].apply(
lambda d: date(d))
expected = expected.set_index(["parcel_id", "inspection_date"])
return expected
def test_one_inspection_one_crime_missing_tract(self):
crimes = [("16Sep2014", "tract567", 3)]
parcels = [("parcelA", "01Dec2014", "tract567")]
population = []
window = datetime.timedelta(days=365)
expected = [("parcelA", date("01Dec2014"), np.nan)]
actual =crime.crimerate_in_aggregation_area(self.make_parcels_df(parcels), self.make_crimerate_df(crimes),
self.make_population_df(population), window)
actual = actual.reset_index()[["parcel_id", "inspection_date", "crime_rate"]].values
self.assert_array_almost_equal(expected, actual)
def test_one_inspection_several_crimes(self):
crimes = [("16Sep2014", "tract567", 3), ("18Oct2014", "tract567", 1)]
parcels = [("parcelA", "01Dec2014", "tract567")]
population = [("tract567", 1234)]
window = datetime.timedelta(days=365)
expected = [("parcelA", date("01Dec2014"), 4 / float(1234))]
actual = crime.crimerate_in_aggregation_area(
self.make_parcels_df(parcels), self.make_crimerate_df(crimes),
self.make_population_df(population), window)
actual = actual.reset_index()[[
"parcel_id", "inspection_date", "crime_rate"
]].values
self.assert_array_almost_equal(expected, actual)
def make_parcels_df(self, parcels):
parcels = pd.DataFrame(
parcels, columns=["parcel_id", "inspection_date", "agg_area"])
parcels["inspection_date"] = parcels["inspection_date"].apply(
lambda d: date(d))
return parcels
def make_expected_df(self, expected):
expected = pd.DataFrame(expected, columns=["parcel_id", "inspection_date", "crime_rate"])
expected["inspection_date"] = expected["inspection_date"].apply(lambda d: date(d))
expected = expected.set_index(["parcel_id", "inspection_date"])
return expected
def make_parcels_df(self, parcels):
parcels = pd.DataFrame(parcels, columns=["parcel_id", "inspection_date", "agg_area"])
parcels["inspection_date"] = parcels["inspection_date"].apply(lambda d: date(d))
return parcels
def make_crimerate_df(self, crimes):
crimes = pd.DataFrame(crimes, columns=["date_reported", "agg_area", "count"])
crimes["year_reported"] = crimes["date_reported"].apply(lambda d: date(d).year)
crimes["month_reported"] = crimes["date_reported"].apply(lambda d: date(d).month)
crimes = crimes.set_index("agg_area")
return crimes[["year_reported", "month_reported", "count"]]