Python Pandas多级索引添加行并执行角度相关运算求助

Hey there! As someone who stumbled through similar Pandas tasks when I was starting out, I’ll walk you through a straightforward, efficient way to tackle this. Let’s turn your original DataFrame into the expanded version with angle-based trigonometric calculations.

Step 1: Break Down the Core Task #

You want to take each row in your original DataFrame, create 360 copies (one for every angle from 0 to 359), then compute new X/Y values using cosine/sine functions that combine the angle with the original X/Y values.

Step 2: Set Up Your Tools & Sample Data #

First, let’s get our imports and sample DataFrame ready (matching your example):

import pandas as pd
import numpy as np

# Your original indexed DataFrame
df = pd.DataFrame({
    'X': [50, 30],
    'Y': [100, 70]
}, index=['Data', 'Data2'])

Step 3: Expand Rows with All Angles #

We’ll first attach a list of all angles (0-359) to every row, then use explode() to turn each list item into its own row. This creates the Cartesian product of your original rows and angles—exactly what you need:

# Add a column containing all angles (0 to 359) to each row
df['Angle'] = [np.arange(360) for _ in range(len(df))]

# Explode the list into individual rows, keeping the original index
expanded_df = df.explode('Angle', ignore_index=False)

# Convert Angle values to integers (explode keeps them as numpy ints initially)
expanded_df['Angle'] = expanded_df['Angle'].astype(int)

After this step, each original row will have 360 entries, each with a unique angle from 0 to 359.

Step 4: Calculate New X/Y with Trigonometric Functions #

Now we’ll compute the updated X/Y values. Let’s use a common formula as an example (adjust this to match your exact requirements):

• new_X = original_X + (some_radius * cos(angle_in_radians))
• new_Y = original_Y + (some_radius * sin(angle_in_radians))

We’ll use vectorized operations (way faster than looping through rows):

# Convert angles from degrees to radians (required for numpy trig functions)
angle_rad = np.radians(expanded_df['Angle'])

# Replace this with your actual radius or variable
radius = 5

# Compute new X and Y values
expanded_df['X'] = expanded_df['X'] + radius * np.cos(angle_rad)
expanded_df['Y'] = expanded_df['Y'] + radius * np.sin(angle_rad)

If you want to keep the original X/Y values alongside the new ones, just rename the columns first (e.g., expanded_df['Original_X'] = expanded_df['X'] before calculating the new X).

Step 5: Match Your Sample’s Index Format (Optional) #

To make the index look like your example (with duplicate index labels hidden), you can mask the repeats when displaying:

# Create a mask for the first occurrence of each index
mask = expanded_df.index.duplicated()

# Set duplicate index labels to empty strings for cleaner display
expanded_df.index = expanded_df.index.where(~mask, '')

This will show the index name only once per original row group, just like your sample output.

Quick Performance Tip #

Always prefer Pandas/Numpy vectorized operations over looping through rows—they’re exponentially faster, especially when you’re generating 360 rows per original entry.

内容的提问来源于stack exchange，提问作者David