Python中移除DataFrame的CHROM列chr字符串的正确方法

你的DataFrame中CHROM列是混合类型（包含整数和字符串），直接用str.replace会导致非字符串类型的元素被转为NaN——因为str系列方法仅对字符串类型生效，非字符串元素会被识别为缺失值。

解决方案一：统一转为字符串后替换，再转回整数 #

先将整列转为字符串类型，确保所有元素都能被字符串处理方法识别，替换完成后再转回整数类型，保持列数据类型一致：

import pandas as pd    
data_as_dict={'CHROM': {232: 1, 233: 1, 234: 1, 10: 'chr15', 11: 'chr15'}, 'POS_GRCh38': {232: 10506158, 233: 109655507, 234: 113903258, 10: '67165147', 11: '67163292'}, 'REF': {232: 'G', 233: 'CAAA', 234: 'G', 10: 'G', 11: 'C'}, 'Effect_allele': {232: 'A', 233: 'C', 234: 'A', 10: 'C', 11: 'T'}, 'Effect_size': {232: 0.1109, 233: 0.0266, 234: 0.0579, 10: 0.2070141693843261, 11: 0.2151113796169455}, 'TYPE': {232: 'Mavaddat_2019_ER_NEG_Breast', 233: 'Mavaddat_2019_ER_NEG_Breast', 234: 'Mavaddat_2019_ER_NEG_Breast', 10: 'THYROID_PGS', 11: 'THYROID_PGS'}, 'Cancer': {232: 'Breast', 233: 'Breast', 234: 'Breast', 10: 'THYROID', 11: 'THYROID'}, 'Significant_YN': {232: 'Y', 233: 'Y', 234: 'Y', 10: 'Y', 11: 'Y'}} 

all_cancers = pd.DataFrame.from_dict(data_as_dict)

# 核心处理代码
all_cancers['CHROM'] = all_cancers['CHROM'].astype(str).str.replace(r'chr', '', regex=True).astype(int)

解决方案二：用apply自定义处理逻辑 #

通过apply遍历每个元素，针对性处理字符串类型的元素，非字符串元素直接保留，最后统一转整数：

def clean_chrom(x):
    if isinstance(x, str):
        return x.replace('chr', '')
    return x

all_cancers['CHROM'] = all_cancers['CHROM'].apply(clean_chrom).astype(int)

两种方法处理后，CHROM列的结果都会变为：

内容的提问来源于stack exchange，提问作者Achal Neupane