如何将DataFrame中B列的分数格式字符串转换为float类型？

Got it, let's figure out how to turn those mixed number strings like '16-1/4' into proper float values in your DataFrame's B column. These are basically whole numbers plus fractions, so we just need to parse each part and do the math. Here are a few reliable approaches depending on your needs:

Method 1: Custom Helper Function (Straightforward & Safe) #

This is my go-to for clarity, especially if you might have edge cases to handle. We'll write a function that breaks down each string, calculates the total value, and can include error handling if needed.

import pandas as pd

def mixed_str_to_float(mixed_str):
    # Split the string into whole number and fraction components
    whole_num, fraction = mixed_str.split('-')
    # Split the fraction into numerator and denominator
    numerator, denominator = fraction.split('/')
    # Compute the total float value
    return int(whole_num) + int(numerator) / int(denominator)

# Apply the function to column B
df['B'] = df['B'].apply(mixed_str_to_float)

If you expect some invalid entries (like strings that don't follow the X-Y/Z format), add a try-except block to avoid crashes:

def mixed_str_to_float_safe(mixed_str):
    try:
        whole_num, fraction = mixed_str.split('-')
        numerator, denominator = fraction.split('/')
        return int(whole_num) + int(numerator) / int(denominator)
    except (ValueError, AttributeError):
        # Return NaN for invalid values to keep the column numeric
        return pd.NA

Method 2: Shortcut with str.replace and eval (For Trusted Data) #

If you're 100% sure your data doesn't contain any malicious or malformed strings, this one-liner is super concise. We just replace the '-' with '+' so the string becomes a valid arithmetic expression, then let eval compute it:

df['B'] = df['B'].str.replace('-', '+').apply(eval)

For example, '16-1/4' turns into '16+1/4', which eval evaluates to 16.25. Quick and dirty, but never use this with untrusted input (since eval executes arbitrary code).

Method 3: Vectorized Operations (Fast for Large Datasets) #

If you're working with a huge DataFrame, vectorized operations will outperform apply by a lot. We'll split the column into parts, convert to numeric types, then compute the total in bulk:

# Split B into whole number and fraction columns
df[['whole', 'frac']] = df['B'].str.split('-', expand=True)
# Split the fraction into numerator and denominator
df[['num', 'den']] = df['frac'].str.split('/', expand=True)

# Convert all temporary columns to integers
df[['whole', 'num', 'den']] = df[['whole', 'num', 'den']].astype(int)

# Calculate the float value for B
df['B'] = df['whole'] + df['num'] / df['den']

# Clean up temporary columns if you don't need them
df = df.drop(['whole', 'frac', 'num', 'den'], axis=1)

This method avoids looping through each row, making it much faster for large datasets.

内容的提问来源于stack exchange，提问作者yeungcase