如何按时间序列索引合并两个Pandas DataFrame？

Hey Joseph, let's break down your two merging problems and get you the result you want:

1. Why pd.concat([FSample, GMSample], axis=1) is stacking rows instead of columns? #

This is a bit unexpected because axis=1 is supposed to merge side-by-side (column-wise)—row stacking is what axis=0 does. Here are the two most likely causes:

• You might have accidentally typed axis=0 instead of axis=1 (easy typo!);
• Your two DataFrames have no overlapping indices, so when you try to merge side-by-side, pandas fills missing rows with NaNs. This can make the result look like the data is stacked vertically, even though it's actually horizontal.

Fix for concat: #

First, make sure your Date field is set as the index for both DataFrames (so pandas can align rows correctly):

# If 'Date' is a column in both DataFrames, set it as the index
FSample = FSample.set_index('Date')
GMSample = GMSample.set_index('Date')

# Now do the column-wise concat
result = pd.concat([FSample, GMSample], axis=1)

This will align rows by the Date index and merge columns side-by-side.

2. Why pd.merge(..., on='Date') throws a 'Date' KeyError? #

This error means one or both of your DataFrames don't have a column named 'Date'. Let's check two things:

• Is Date actually the index of one/both DataFrames instead of a column? Use FSample.columns and GMSample.columns to list all column names and confirm;
• Did you misspell the column name? (e.g., lowercase date, or Date_ with a suffix)

Fix for merge: #

Case 1: Date is the index (not a column)

Use left_index=True and right_index=True to merge on the index instead of a column:

# Inner join using the Date index
result = pd.merge(FSample, GMSample, how='inner', left_index=True, right_index=True)

Case 2: Date is a column but misnamed

Fix the column name first, then merge:

# Example: if GMSample uses 'date' instead of 'Date'
GMSample.rename(columns={'date': 'Date'}, inplace=True)

# Now merge on the corrected column name
result = pd.merge(FSample, GMSample, how='inner', on='Date')

Quick Debug Tip #

To avoid confusion, print out basic info about your DataFrames first:

print("FSample Columns:", FSample.columns.tolist())
print("FSample Index Name:", FSample.index.name)
print("\nGMSample Columns:", GMSample.columns.tolist())
print("GMSample Index Name:", GMSample.index.name)

This will instantly tell you if Date is a column or index, and if there are spelling mismatches.

内容的提问来源于stack exchange，提问作者Joseph U.