def apply_sequencing(self,
d=0.1,
n=40,
distance=False,
fixed=False,
dwelling=False,
enumerator=False,
tuned=False,
enumerator_dist=False):
print('d: ', d)
sequences_dfs = []
if distance:
dwellings_dist = tuned if tuned else self.get_dwellings_dist_seq(d)
sequences_dfs.append(dwellings_dist)
if fixed:
sequences_dfs.append(self.get_dwellings_fixed_seq(n))
if dwelling:
sequences_dfs.append(self.get_dwellings_dwellings_seq())
if enumerator:
sequences_dfs.append(self.get_enum_seq())
if enumerator_dist:
dwellings_dist = self.get_enum_seq(all=True)
dwellings_dist.dropna(subset=[self.block_col], inplace=True)
dwellings_dist = dwellings_dist.groupby(
[self.ward_col, "enum_seq"],
as_index=False).apply(lambda x: sequences.col_for_seq(
x, X=self.x_col, Y=self.y_col))
dwellings_dist = dwellings_dist.groupby(
[self.ward_col, "enum_seq"],
as_index=False).apply(lambda x: sequences.get_dist_seq(x, d))
dwellings_dist.rename(columns={
"sequence_id": "enum_dist_id",
"sequence_order_enum": "enum_dist_order",
"dist": "enum_dist",
"sequence_len": "enum_sequence_len"
},
inplace=True)
sequences_dfs.append(dwellings_dist.loc[:, [
self.ward_col, self.dwelling_col, "enum_dist_id",
"enum_dist_order", "enum_dist", "enum_sequence_len"
]].copy())
self.df = reduce(
lambda x, y: pd.merge(
x, y, how="left", on=[self.ward_col, self.dwelling_col]),
sequences_dfs, self.data)
## fill within for distance sequence
# if distance:
# self.df = self.df.groupby(self.ward_col, as_index=False).apply(lambda x: self.fill_within(x, self.dwelling_col, 'sequence_id'))
if enumerator_dist:
# self.df = self.df.groupby(self.ward_col, as_index=False).apply(lambda x: self.fill_within(x, self.dwelling_col, 'enum_dist_id'))
self.df = self.df.dropna(subset=[self.pagenum])
def sequence_datasets(census_1850, census_1880):
census_1850 = census_1850.dropna(subset=["CENSUS_DWELLING_NUM"]).copy()
dwellings_1850 = census_1850.groupby(["WARD_NUM", "CENSUS_DWELLING_NUM"],
as_index=False).first()
dwellings_1850 = dwellings_1850.dropna(subset=["CD_ADDRESS"]).copy()
dwellings_1880 = census_1880.drop_duplicates(
subset=["CENSUS_ADDRESS"]).reset_index(drop=True).copy()
sequences.col_for_seq(dwellings_1850, "CD_X", "CD_Y")
sequences.col_for_seq(dwellings_1880, "POINT_X", "POINT_Y")
dwellings_1850 = sequences.get_dist_seq(dwellings_1850, 0.15)[2]
dwellings_1880 = sequences.get_dist_seq(dwellings_1880, 0.15)[2]
dwellings_1850 = dwellings_1850.groupby("sequence_id").apply(
sequences.sequence_order)
dwellings_1880 = dwellings_1880.groupby("sequence_id").apply(
sequences.sequence_order)
#Not super sure what's happening here, come back and check this
dwellings_1850 = dwellings_1850.groupby(
["WARD_NUM", "CENSUS_DWELLING_NUM"], as_index=False).first()
dwellings_1880 = dwellings_1880.drop_duplicates(
subset=["CENSUS_ADDRESS"]).reset_index(drop=True).copy()
census_1880_model = dataprocessing.dwellings_to_all(
census_1880, dwellings_1880,
["CENSUS_MATCH_ADDR", "sequence_id", "sequence_order", "num_between"],
["CENSUS_MATCH_ADDR"])
census_1850_model = dataprocessing.dwellings_to_all(
census_1850, dwellings_1850, [
"WARD_NUM", "CENSUS_DWELLING_NUM", "sequence_id", "sequence_order",
"num_between", "sequence_order_enum"
], ["WARD_NUM", "CENSUS_DWELLING_NUM"])
dataprocessing.create_street_house(dwellings_1880, "CENSUS_ADDRESS")
dataprocessing.create_street_house(dwellings_1850, "CD_ADDRESS")
dataprocessing.create_street_house(census_1880_model, "CENSUS_ADDRESS")
dataprocessing.create_street_house(census_1850_model, "CD_ADDRESS")
return [
dwellings_1850, dwellings_1880, census_1850_model, census_1880_model
]
def get_dwellings_dist_seq(self, d=0.1):
dwellings = self.get_dwellings()
dwellings.dropna(subset=[self.block_col], inplace=True)
dwellings_cols = dwellings.groupby(
self.ward_col, as_index=False).apply(
lambda x: sequences.col_for_seq(x, X=self.x_col, Y=self.y_col))
dwellings_cols = dwellings_cols.groupby(
self.ward_col,
as_index=False).apply(lambda x: sequences.get_dist_seq(x, d))
return dwellings_cols.loc[:, [
self.ward_col, self.dwelling_col, "sequence_id", "num_between",
"sequence_order_enum", "dist", "sequence_len"
]].copy()