如何基于现有列条件创建DataFrame新列？代码执行异常求助

Hey there! Let's break down why your apply statement is only hitting the ELSE branch—you're checking for NaN values the wrong way.

The Root Problem #

In Pandas, missing values (NaN) aren't strings like "NAN" or "NaN"—they're special floating-point values (or pd.NA for nullable dtypes). When you write x['Eval'] == 'NAN', you're comparing a numeric/nullable value to a string, which will always return False. That's why your code never triggers the IF branch!

Correct Solutions #

Here are a few ways to fix this, from direct fixes to more efficient alternatives:

1. Fix the Lambda with isna()

Use Pandas' built-in isna() function to properly detect missing values:

import pandas as pd

dataFrame01['final'] = dataFrame01.apply(lambda x: x['Name'] if pd.isna(x['Eval']) else x['Eval'], axis=1)

Or you can call the isna() method directly on the column value:

dataFrame01['final'] = dataFrame01.apply(lambda x: x['Name'] if x['Eval'].isna() else x['Eval'], axis=1)

2. Faster Vectorized Alternatives (Recommended!)

Using apply(axis=1) can be slow for large datasets. Instead, use vectorized operations which are much more efficient:

Option A: np.where

import numpy as np

dataFrame01['final'] = np.where(dataFrame01['Eval'].isna(), dataFrame01['Name'], dataFrame01['Eval'])

Option B: fillna (Simplest!)

Since you're replacing missing values in Eval with values from Name, fillna is the most concise approach:

dataFrame01['final'] = dataFrame01['Eval'].fillna(dataFrame01['Name'])

Edge Case Check #

If your data actually has string values like "NAN" (not true missing values), then you'd use x['Eval'] == 'NAN'—but this is rare. Double-check your Eval column's dtype with dataFrame01['Eval'].dtype to confirm. If it's object dtype, you might have string-based "missing" values instead of proper NaNs.

内容的提问来源于stack exchange，提问作者Mohd Faheem