5 Best Ways to Get Day of the Month From a Period with Python Pandas

by

💡 Problem Formulation: When working with time series data in Python Pandas, you might need to extract the day of the month that a certain period falls on. For example, given a Period object for August 2020, you might need to determine which day of the month the first day of August falls on, which would be ‘1’.

Method 1: Using Period.day Attribute

Python Pandas provides a Period object that represents time intervals. The Period object has a day attribute which can be used to obtain the day of the month directly. This is the most intuitive method for those already familiar with Period objects in Pandas.

Here’s an example:

import pandas as pd

# Create a period for the month of August 2020
period = pd.Period('2020-08')

# Get the day of the month
day_of_month = period.day
print("The day of the month for the period is:", day_of_month)

Output:

The day of the month for the period is: 1

This code snippet creates a period for the month of August 2020 and then simply accesses the day attribute of the Period object to print the day of the month the period starts on, which is 1.

Method 2: Using to_timestamp() and day Attribute

Another approach to retrieve the day of the month is to convert the Period to a Timestamp with to_timestamp() method, then use the day attribute of the resulting Timestamp. This is especially useful if the period is not at daily frequency and you need the day of the specific start or end time.

Here’s an example:

import pandas as pd

# Create a period for the first quarter of 2020
period = pd.Period('2020Q1')

# Convert to timestamp, i.e., the start of the period
timestamp = period.to_timestamp()

# Get the day of the month
day_of_month = timestamp.day
print("The day of the month for the period is:", day_of_month)

Output:

The day of the month for the period is: 1

In this code snippet, we create a Period for the first quarter of 2020. We then convert it to a Timestamp to get the exact starting point of the period, after which we access the day attribute. The output is 1, indicating the period starts on the first day of January.

Method 3: Using Period.days_in_month

Suppose you’re interested in finding the day of the month on which the period ends rather than begins. For this, the Period.days_in_month attribute can come in handy, as it returns the number of days in that month — effectively telling you what the last day of the period’s month is.

Here’s an example:

import pandas as pd

# Create a period for February 2020
period = pd.Period('2020-02')

# Get the last day of the month for a period
last_day_of_month = period.days_in_month
print("The last day of the month for the period is:", last_day_of_month)

Output:

The last day of the month for the period is: 29

This snippet showcases how to get the last day of February 2020 using the days_in_month attribute, which tells us there are 29 days due to it being a leap year.

Method 4: Using Period.end_time Property

Aliasing somewhat to the previous method, but instead of getting the number of days, the Period.end_time property is used to get the exact moment a period ends. The day can then be extracted similarly to the previous methods.

Here’s an example:

import pandas as pd

# Create a period for March 2020
period = pd.Period('2020-03')

# Get the end time of the period
end_time = period.end_time

# Extract the day from the end time
last_day_of_month = end_time.day
print("The last day of the month for the period is:", last_day_of_month)

Output:

The last day of the month for the period is: 31

By getting the end time (which is a Timestamp) of the period using end_time, and then querying its day attribute, we find out March 2020 ended on the 31st.

Bonus One-Liner Method 5: Using a Lambda Function

For those who love simplifying scripts to one-liners, using a lambda function to compact Method 1 or Method 2 can be appealing. This is useful for applying the technique to an entire Series or DataFrame column quickly.

Here’s an example:

import pandas as pd

# Create a Series of periods
periods = pd.Series(pd.period_range('2021-01', periods=3, freq='M'))

# Get the day of the month for each period using a lambda function
days_of_month = periods.apply(lambda x: x.day)
print(days_of_month)

Output:

0    31
1    28
2    31
dtype: int64

This one-liner uses apply() on a Pandas Series of Period objects, with a lambda function extracting the final day of each month using the day attribute. It efficiently applies the extraction over the entire series.

Summary/Discussion

  • Method 1: Using Period.day. Straightforward and simple, best for single day-of-month queries. Not suitable for ranges or multiple periods.
  • Method 2: Using to_timestamp() and day. Highly versatile, allows for a clear transition between Periods and Timestamps. More steps than necessary for simple queries.
  • Method 3: Using Period.days_in_month. Ideal for finding the out-of-box end of a period’s month. Doesn’t give information about the start of the period.
  • Method 4: Using Period.end_time Property. Provides precise end time of the Period, more detailed than Method 3. Slightly redundant if only the day is needed.
  • Method 5: Using a Lambda Function. Extremely compact and perfect for applying over Series or DataFrames, but less readable for beginners.