使用Pandas将DataFrame中world_rank列转换为float类型的问题求助

It looks like your issue stems from two key problems: your initial replace was targeting all hyphens (including those in range values like 601-800), and your custom function didn’t handle NaN values or edge-case formats properly. Let’s walk through a robust solution step by step.

Step 1: Fix Missing Value Handling #

First, don’t replace all hyphens with NaN—that breaks your range values. Instead, only replace standalone hyphens (which represent missing ranks) using regex:

import pandas as pd
import numpy as np
import re

# Replace ONLY standalone "-" with NaN (preserves hyphens in ranges like 601-800)
times_df['world_rank'] = times_df['world_rank'].replace(r'^-$', np.nan, regex=True)

Step 2: Robust Midpoint Calculation Function #

Your original function failed because it tried to process NaN values (which aren’t strings) and didn’t account for edge cases like extra whitespace or invalid characters. Here’s an improved version:

def mid_rank(rank_entry):
    # Handle NaN values first to avoid string operation errors
    if pd.isna(rank_entry):
        return np.nan
    
    # Convert to string, remove equal signs, and strip extra whitespace
    rank_clean = re.sub(r'=', '', str(rank_entry)).strip()
    
    # Split by hyphen to process range values
    rank_parts = rank_clean.split('-')
    
    # Try to convert parts to floats and calculate midpoint
    try:
        # Clean each part and filter out empty strings
        numeric_parts = [float(part.strip()) for part in rank_parts if part.strip()]
        
        # Return average if we have valid numeric values
        if numeric_parts:
            return sum(numeric_parts) / len(numeric_parts)
        else:
            return np.nan
    except ValueError:
        # Catch any unconvertible formats (e.g., non-numeric characters)
        return np.nan

Step 3: Apply the Function #

Now you can safely generate your cleaned numeric rank column:

times_df['world_rank_tidy'] = times_df['world_rank'].apply(mid_rank)

Why This Works: #

• Preserves ranges: We only replace standalone hyphens, so values like 601-800 stay intact to split and calculate midpoints.
• Handles NaNs: The function checks for NaN first to avoid errors when processing missing values.
• Edge case resilience: The try-except block and whitespace handling ensure even messy formats (like = 901-1000) get processed correctly.

内容的提问来源于stack exchange，提问作者SQL Learner 1