如何将DataFrame中%m-%d格式的日期字符串转换为dayofyear格式？

Hey there! Let's figure out how to convert those '%m-%d' date strings to day-of-year values in your pandas DataFrame. It's actually pretty straightforward with pandas' built-in datetime tools—here's a step-by-step breakdown:

Step 1: Convert the string column to datetime objects #

First, we need to turn those '%m-%d' strings into proper datetime entries. Since your dates don't include a year, pandas will automatically use the current year by default. If you want to use a specific year (like handling leap years for February 29), you can explicitly set it with the year parameter.

Example code to get started:

import pandas as pd

# Your sample DataFrame
df = pd.DataFrame({'month_day': ['01-01', '02-28', '03-15', '12-31', '02-29']})

# Option 1: Use current year (default behavior)
df['date'] = pd.to_datetime(df['month_day'], format='%m-%d')

# Option 2: Specify a fixed year (e.g., 2024, a leap year)
df['date_fixed'] = pd.to_datetime(df['month_day'], format='%m-%d', year=2024)

Step 2: Extract the day-of-year value #

Once you have datetime objects, you can use the .dt.dayofyear attribute to get the day number (1 to 365/366):

# Get dayofyear for both date columns
df['dayofyear_current'] = df['date'].dt.dayofyear
df['dayofyear_fixed'] = df['date_fixed'].dt.dayofyear

Handling edge cases: Invalid dates #

If your data has dates like '02-29' and you're using a non-leap year (e.g., 2023), pandas will throw an error by default. To handle this gracefully, add errors='coerce' to convert invalid dates to NaT (Not a Time), which will result in NaN for the day-of-year value:

# Safe conversion with error handling
df['date_safe'] = pd.to_datetime(df['month_day'], format='%m-%d', year=2023, errors='coerce')
df['dayofyear_safe'] = df['date_safe'].dt.dayofyear

Example output #

If you print the final DataFrame, you'll see something like this:

Notice how '03-15' has different day-of-year values in 2023 (non-leap) vs 2024 (leap)—that's why specifying a fixed year can be important depending on your use case!

内容的提问来源于stack exchange，提问作者Rishi