如何用Python基于列表变量对DataFrame执行差值计算并新增列？

Got it, let's work through this problem step by step. Here's a straightforward, scalable way to compute those time-based difference columns in your pandas DataFrame:

First, we'll loop through each base variable (OUT and IN), then calculate the difference between consecutive time-suffixed columns (like 3M minus 6M, 6M minus 9M, etc.) and assign these results to new, clearly named columns.

Full Code Implementation #

import pandas as pd

# Your sample input DataFrame
Input_df = pd.DataFrame({
    'ID': ['A', 'B', 'C', 'D'],
    'OUT_3M': [2, 3, 2, 3],
    'OUT_6M': [3, 3, 3, 3],
    'OUT_9M': [4, 5, 6, 7],
    'OUT_15M': [6, 7, 6, 7],
    'IN_3M': [2, 3, 2, 3],
    'IN_6M': [3, 3, 3, 3],
    'IN_9M': [4, 5, 6, 7],
    'IN_15M': [6, 7, 6, 7]
})

# Define your base variables and time suffix order
target_vars = ['OUT', 'IN']
time_periods = ['3M', '6M', '9M', '15M']

# Calculate differences for each variable and time pair
for var in target_vars:
    # Iterate over consecutive time periods
    for idx in range(len(time_periods) - 1):
        current_period = time_periods[idx]
        next_period = time_periods[idx + 1]
        # Create new column name (e.g., OUT_3M-6M)
        new_col_name = f"{var}_{current_period}-{next_period}"
        # Compute the difference and add to the DataFrame
        Input_df[new_col_name] = Input_df[f"{var}_{current_period}"] - Input_df[f"{var}_{next_period}"]

# Preview the updated DataFrame
print(Input_df)

Breakdown of the Approach #

• Define time order: We list the time suffixes in the sequence we need to compare (3M → 6M → 9M → 15M) to ensure we calculate the correct consecutive differences.
• Loop through base variables: For each variable (OUT and IN), we process all its time-suffixed columns without repeating code.
• Dynamic column naming: The new column names follow your requested format (e.g., Out_3M-6M) so results are easy to identify.
• Efficient calculation: We directly compute and assign the differences to the original DataFrame, avoiding unnecessary copies of data.

When you run this code, your Input_df will include all the new difference columns. For example, the OUT_3M-6M column will show values like -1, 0, -1, 0 for rows A-D respectively, matching the sample data you provided.

内容的提问来源于stack exchange，提问作者Zhoe