class Data:
def __init__(self, *args, **kwargs):
self.power = ""
# get data and merge on 'tds'
if "Electricity" in args:
self.power = "Electricity"
self.data = Electricity().get_data(**kwargs)
else:
self.power = "Water"
self.data = Water().get_data(**kwargs)
self.coords = Coords().get_all_data()
# cleanup the data
self.merge_coords()
self.add_year_column()
self.set_dtypes()
# compute df of mean consumption by (tds x year)
self.tds_by_year = None
self.get_by_year_data()
# add other pre-computed dataframes here
# ...
self.tds_sum_year = None
self.get_aggregate_data()
# Merging coordinates with other datasets on TDS value
def merge_coords(self):
"""
Merge the coords into (lat,long) into the data
"""
self.data = self.data.merge(self.coords, on="tds")
# Setting the datatype of the merged dataset
def set_dtypes(self):
"""
Set the datatypes to intended types
"""
self.data.latitude = self.data.latitude.astype(float)
self.data.longitude = self.data.longitude.astype(float)
self.data.year = self.data.year.astype(int)
if self.power == "Electricity":
self.data.consumption_kwh = self.data.consumption_kwh.astype(float)
else:
self.data.consumption_hcf = self.data.consumption_hcf.astype(float)
# Adding a year column from the revenue month column
def add_year_column(self):
"""
Create a new column with the year from revenue_month
"""
years = self.data["revenue_month"].apply(lambda x: int(x.split("-")[0]))
self.data["year"] = years
# Getting the mean consumption for each year
def get_by_year_data(self):
"""
Get difference in the mean consumption between years.
"""
if self.power == "Electricity":
self.tds_by_year = self.data.groupby(["tds", "year"]).consumption_kwh.mean()
else:
self.tds_by_year = self.data.groupby(["tds", "year"]).consumption_hcf.mean()
# Getting the total consumption for each year
def get_aggregate_data(self):
"""
Get the total consumption for each year
"""
if self.power == "Electricity":
self.tds_sum_year = self.data.groupby(["tds", "year"]).consumption_kwh.sum()
else:
self.tds_sum_year = self.data.groupby(["tds", "year"]).consumption_hcf.sum()
