如何在Pandas中展平多级索引，使所有列与Date同级并将Date设为索引？

Let's break down what's going wrong in your code and fix it step by step:

Two Key Issues in Your Current Code #

1. You didn't assign the result of set_index('Date') back to your DataFrame
   Pandas methods like set_index() return a new DataFrame by default—they don't modify the original object in place. So your data variable stayed unchanged after calling this line.
2. Using to_records() is unnecessary
   Converting to a structured record array and back to a DataFrame is a roundabout way to handle the stacked data; we can clean this up with more straightforward pandas operations.

Corrected Code #

First, make sure you import pandas (you missed this in your original snippet):

import pandas as pd
from pandas_datareader import data

Then process the data with cleaner, explicit steps:

stocks = ['GOOG', 'AAPL','MSFT']
# Fetch data from Yahoo Finance
df = data.DataReader(stocks,'yahoo')

# Stack the second level of column index (stock symbols) into row index
df_stacked = df.stack(level=1)
# Convert the multi-level row index into regular columns
df_stacked = df_stacked.reset_index()
# Rename the auto-generated 'level_1' column to 'Symbols'
df_stacked = df_stacked.rename(columns={'level_1': 'Symbols'})
# Set Date as the index, and assign back to update the DataFrame
df_final = df_stacked.set_index('Date')

Or use chained operations for brevity:

df_final = (data.DataReader(stocks,'yahoo')
            .stack(level=1)
            .reset_index()
            .rename(columns={'level_1': 'Symbols'})
            .set_index('Date'))

Verify the Result #

Check the columns and index to confirm everything is as expected:

print(df_final.columns)
# Output: Index(['Symbols', 'Adj Close', 'Close', 'High', 'Low', 'Open', 'Volume'], dtype='object')

print(df_final.index.name)
# Output: 'Date'

How This Works #

• stack(level=1) moves the stock symbols (second level of the original column index) into the row index, creating a multi-level index of (Date, Symbols).
• reset_index() converts both levels of the row index into regular columns (Date and the auto-named level_1 for symbols).
• rename() fixes the default column name for symbols to match your desired structure.
• set_index('Date') makes Date the primary index, and assigning the result ensures your DataFrame reflects this change.

内容的提问来源于stack exchange，提问作者Alexander Gavrailov