Python Pandas - Create a PeriodIndex and get the days of the week

To create a PeriodIndex, use the pandas.PeriodIndex() method. Get the days of the week using the PeriodIndex.dayofweek property, which returns Monday=0, Tuesday=1 ... Sunday=6.

What is PeriodIndex?

PeriodIndex is an immutable ndarray holding ordinal values indicating regular periods in time. It's useful for time series analysis where you need to represent fixed-frequency date periods ?

Creating a PeriodIndex

First, import pandas and create a PeriodIndex with specific dates ?

import pandas as pd

# Create a PeriodIndex object with daily frequency
periodIndex = pd.PeriodIndex(['2018-07-25', '2019-10-30', '2020-11-20',
                             '2021-09-15', '2022-03-12', '2023-06-18'], freq="D")

print("PeriodIndex...")
print(periodIndex)

PeriodIndex...
PeriodIndex(['2018-07-25', '2019-10-30', '2020-11-20', '2021-09-15', '2022-03-12', '2023-06-18'], dtype='period[D]')

Getting Days of the Week

Use the dayofweek property to extract the day of the week as integers ?

import pandas as pd

periodIndex = pd.PeriodIndex(['2018-07-25', '2019-10-30', '2020-11-20',
                             '2021-09-15', '2022-03-12', '2023-06-18'], freq="D")

# Get days of the week (Monday=0, Tuesday=1, ..., Sunday=6)
days_of_week = periodIndex.dayofweek
print("Days of the week from the PeriodIndex...")
print(days_of_week)

Days of the week from the PeriodIndex...
Index([2, 2, 4, 2, 5, 6], dtype='int64')

Complete Example

Here's a comprehensive example showing PeriodIndex creation and various properties ?

import pandas as pd

# Create a PeriodIndex object with daily frequency
periodIndex = pd.PeriodIndex(['2018-07-25', '2019-10-30', '2020-11-20',
                             '2021-09-15', '2022-03-12', '2023-06-18'], freq="D")

# Display PeriodIndex object
print("PeriodIndex...")
print(periodIndex)

# Display frequency
print("\nPeriodIndex frequency...")
print(periodIndex.freq)

# Display day numbers
print("\nThe day numbers from the PeriodIndex...")
print(periodIndex.day)

# Display days of the week
print("\nDays of the week from the PeriodIndex...")
print(periodIndex.dayofweek)

PeriodIndex...
PeriodIndex(['2018-07-25', '2019-10-30', '2020-11-20', '2021-09-15', '2022-03-12', '2023-06-18'], dtype='period[D]')

PeriodIndex frequency...
<Day>

The day numbers from the PeriodIndex...
Index([25, 30, 20, 15, 12, 18], dtype='int64')

Days of the week from the PeriodIndex...
Index([2, 2, 4, 2, 5, 6], dtype='int64')

Understanding the Output

The dayofweek values correspond to:

• 0 = Monday
• 1 = Tuesday
• 2 = Wednesday
• 3 = Thursday
• 4 = Friday
• 5 = Saturday
• 6 = Sunday

So our dates [2, 2, 4, 2, 5, 6] represent Wednesday, Wednesday, Friday, Wednesday, Saturday, and Sunday respectively.

Conclusion

Use pd.PeriodIndex() to create time period objects and dayofweek property to extract weekday information as integers from 0-6. This is useful for time series analysis and filtering data by weekdays.