CODEWITHSUNDEEP
pythonpandasdataframeisinpython-holidays
Gonçalo Peres
Gonçalo Peres

Reputation: 13622

isin holidays only recognizing the first hour

I have created a class with the holidays in Spain

class SpainBusinessCalendar(AbstractHolidayCalendar):
   rules = [
     #Spain - If one holiday is on a Sunday, each Autonomous Community can change it to a Monday.
     Holiday('Año Nuevo', month=1, day=1, observance=sunday_to_monday),
     Holiday('Epifanía del Señor', month=1, day=6, observance=sunday_to_monday),
     Holiday('Viernes Santo', month=1, day=1, offset=[Easter(), Day(-2)]),
     Holiday('Día del Trabajador', month=5, day=1, observance=sunday_to_monday),
     Holiday('Asunción de la Virgen', month=8, day=15, observance=sunday_to_monday),
     Holiday('Día de la Hispanidad', month=10, day=12, observance=sunday_to_monday),
     Holiday('Todos los Santos', month=11, day=1, observance=sunday_to_monday),
     Holiday('Día Constitución', month=12, day=6, observance=sunday_to_monday),
     Holiday('Inmaculada Concepción', month=12, day=8, observance=sunday_to_monday),        
     Holiday('Navidad', month=12, day=25, observance=sunday_to_monday)
   ]

Then I generated an with the size equals to the Date column in my dataframe

cal = SpainBusinessCalendar()
holidays = cal.holidays(start=df['Date'].min(), end=df['Date'].max())

Which gives the following

enter image description here

In order to generate a df column with the holidays, resulting from the values in the column "Date", I have done

df['Feriado'] = df['Date'].isin(holidays).astype(int)

However, as one can guess from the image of the holidays output, if one is working with hourly data, which is the case, it will only pick up as holiday the first hour (with the time 00:00).

How should I proceed in order to, in the analysis of holidays, the hour is ignored and for a specific holiday date, assign the respective value.

Edit

Both

holidays = cal.holidays(start=df['Data'].dt.date.min(), end=df['Data'].dt.date.max())

and

holidays = cal.holidays(start=df['Data'].dt.floor('d').min(), end=df['Data'].dt.floor('d').max())

Give the same output as the image above.

Upvotes: 0

Views: 199

Answers (1)

Gonçalo Peres
Gonçalo Peres

Reputation: 13622

Created a column with the Date without the time component

df['Date_notime'] = df['Data'].dt.floor('d')

Then generated the holidays from that column

holidays = cal.holidays(start=df['Date_notime'].dt.date.min(), end=df['Date_notime'].dt.date.max())

And like then

df['Feriado'] = df['Date_notime'].isin(holidays).astype(int)

And at the end dropped the columns that I didn't want.

Upvotes: 1

Related Questions