pandas中的遍历方式速度对比

对一个20667行的xlsx文件进行遍历测试

import pandas as pd

# 定义一个计算执行时间的函数作装饰器，传入参数为装饰的函数或方法
def print_execute_time(func):
    from time import time

    # 定义嵌套函数，用来打印出装饰的函数的执行时间
    def wrapper(*args, **kwargs):
        # 定义开始时间和结束时间，将func夹在中间执行，取得其返回值
        start = time()
        func_return = func(*args, **kwargs)
        end = time()
        # 打印方法名称和其执行时间
        print(f'{func.__name__}() execute time: {end - start}s')
        # 返回func的返回值
        return func_return

    # 返回嵌套的函数
    return wrapper

file_path = r"D:\git\xxxx\dev\pd-xxx1.2\合并.xlsx"
data = pd.read_excel(file_path,sheet_name="xxxx",engine='openpyxl')
# 空值处理
df = data.where(data.notnull(),None)

@print_execute_time
def iterrows():
    for index, row in df.iterrows():
        # print(index," = ",row['机号'])
        pass

@print_execute_time
def itertuples():
    for row in df.itertuples():
        # print(row['机号'])
        pass

@print_execute_time
def iteritems():
    for index, row in df.iteritems():
        # print(index," = ",row['机号'])
        pass

@print_execute_time
def index():
    for i in df.index:
        # print(i," = ",df['机号'].at[i])
        pass

if __name__ == '__main__':
    print('begining ...')
    print(iterrows(),itertuples(),iteritems(),index())
    print('Done !')

测试结果

begining ...
iterrows() execute time: 2.003657817840576s
itertuples() execute time: 0.04618692398071289s
iteritems() execute time: 0.0009987354278564453s
index() execute time: 0.0029909610748291016s
Done !

iterrows() execute time: 2.2464449405670166s
itertuples() execute time: 0.08178043365478516s
iteritems() execute time: 0.000997781753540039s
index() execute time: 0.0059833526611328125s

因此从效率上考虑,优先采用iteritems或index来进行遍历数据