如何用Pandas计算CSV中位数？新手遇KeyError求助

Hey there! Let's break down why you're running into that KeyError and get your median calculation working right away.

Why the KeyError Happens #

Your line median = census[0].median() is trying to access a column named "0" in your DataFrame—but your CSV almost certainly doesn't have a column with that exact name. In Pandas, the square bracket [] operator looks up columns by their label/name, not their position index. Since there's no column called "0", you get the KeyError.

Step-by-Step Fixes & Correct Methods #

First, let's confirm your CSV's structure so you know what you're working with. Add these lines right after loading the CSV to see your column names and sample data:

print("Your CSV columns:", census.columns)
print("\nSample data from your CSV:\n", census.head())

Once you know your column names or positions, use one of these approaches:

1. Calculate median for a specific column (by name, recommended)
   If you want the median of a column like population or age, reference it directly by its name:

   # Replace "your_column_name" with the actual name from census.columns
   median = census["your_column_name"].median()
   

2. Calculate median for a column by position
   If you want the first column (index 0) regardless of its name, use iloc (integer location) to target columns by their position:

   # : means all rows, 0 means the first column
   median = census.iloc[:, 0].median()
   

3. Calculate median for all numeric columns at once
   If you want medians for every numeric column in your CSV, use the DataFrame's built-in median() method with numeric_only=True to skip non-numeric columns:

   all_medians = census.median(numeric_only=True)
   print("Medians for all numeric columns:\n", all_medians)
   

Full Working Example #

Here's how your code might look after fixing it (using the position-based approach as an example):

import pandas
census = pandas.read_csv("census(1).csv")

# Verify your data structure first
print("CSV Columns:", census.columns)
print("\nFirst 5 rows:\n", census.head())

# Calculate median for the first column
median = census.iloc[:, 0].median()
print("\nMedian of the first column:", median)

If you're still stuck, sharing a few rows of your CSV (including the header row) would help debug further—sometimes numeric columns get incorrectly parsed as strings, which can also cause issues with median calculations.

内容的提问来源于stack exchange，提问作者Alex