def process_location(df: pd.DataFrame,
monotonic_check: bool = True) -> pd.DataFrame:
# Only report up to previous day to avoid partial reporting
df = df.assign(date=pd.to_datetime(df.date, dayfirst=True))
df = df[df.date.dt.date < datetime.now().date()]
# Default columns for second doses
if "people_vaccinated" not in df:
df = df.assign(people_vaccinated=pd.NA)
df.people_vaccinated = df.people_vaccinated.astype("Int64")
if "people_fully_vaccinated" not in df:
df = df.assign(people_fully_vaccinated=pd.NA)
df.people_fully_vaccinated = df.people_fully_vaccinated.astype("Int64")
# Avoid decimals
cols = [
"total_vaccinations", "people_vaccinated", "people_fully_vaccinated"
]
df[cols] = df[cols].astype(float).astype("Int64").fillna(pd.NA)
# Order columns and rows
usecols = [
"location", "date", "vaccine", "source_url", "total_vaccinations",
"people_vaccinated", "people_fully_vaccinated"
]
df = df[usecols]
df = df.sort_values(by="date")
# Sanity checks
country_df_sanity_checks(df, monotonic_check=monotonic_check)
# Strip
df = df.applymap(lambda x: x.strip() if isinstance(x, str) else x)
# Date format
df = df.assign(date=df.date.dt.strftime("%Y-%m-%d"))
return df
def read_csv_and_check(filepath):
# Read
try:
df = pd.read_csv(
filepath,
# parse_dates=["date"],
# dayfirst=True
)
except Exception:
raise ValueError(f"Check the spreadsheet corresponding to {filepath}")
location = df.loc[:, "location"].unique()
# Date check
try:
df = df.assign(date=pd.to_datetime(df["date"], format="%Y-%m-%d"))
except Exception:
raise ValueError(f"{location} -- Invalid date format! Should be %Y-%m-%d. Check {df.date}.")
if not df.date.is_monotonic:
raise ValueError(f"{location} -- Check that date field is monotonically increasing.")
# Checks
country_df_sanity_checks(df)
return df