Pandas数据列标记函数出现KeyError问题求助

场景描述 #

从Excel读取Pandas DataFrame后，原本处理单列的代码运行正常：

tmd.iloc[:, 10] = tmd.iloc[:, 10].fillna(tmd.iloc[:, 10].mean())
tmd.iloc[tmd.iloc[:, 10] == 0, 10] = tmd.iloc[:, 10].mean()
tmd.iloc[:, 10] = tmd.iloc[:, 10].where(~((tmd.iloc[:, 10] > 0) & (pd.Series.abs(tmd.iloc[:, 10].diff()) > 30)), tmd.iloc[:, 10].mean())

但为了处理多列编写checkFlags_current函数后，执行 tmd_flags['Load Current(A)'] = checkFlags_current(tmd.iloc[:, 10])时出现KeyError：

def checkFlags_current(var):
    """
        This function calculates,

            Input(s):
            - var1: Current in Ampere

            Returns:
            - flags
    """
    rows = len(var)
    flags = np.zeros((rows, 1))

    for i in range(0, rows):
        if (var[i] > 0 & (abs(var[i+1] - var[i]) > 30)):
            flags[i] = 1
        elif (pd.isnull(var[i])):
            flags[i] = 3
        elif (var[i] == 0):
            flags[i] = 2
        else:
            flags[i] = 0
    flags = list(itertools.chain(*flags))

    return flags

错误信息：

File "D:\AssetManager\Scripts\req_functions.py", line 1215, in checkFlags_current
    if (var[i] > 0 & (abs(var[i+1] - var[i]) > 30)):
  File "C:\Users\jadha\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\series.py", line 958, in __getitem__
    return self._get_value(key)
  File "C:\Users\jadha\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\series.py", line 1069, in _get_value
    loc = self.index.get_loc(label)
  File "C:\Users\jadha\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\pandas\core\indexes\range.py", line 387, in get_loc
    raise KeyError(key) from err
KeyError: 32073

错误原因 #

1. 循环越界：当i遍历到最后一行（索引为rows-1）时，i+1超出了Series的索引范围，导致无法找到对应标签。
2. 运算符优先级错误：var[i] > 0 & (...)中，位运算符&的优先级高于比较运算符>，会先计算0 & (...)，导致逻辑判断错误。
3. 索引访问方式错误：用var[i]是按标签索引，而非位置索引，如果Series的索引不是连续整数（或默认RangeIndex），会出现KeyError，应该用iloc按位置访问。

修复方案 #

方案1：修复循环逻辑 #

import numpy as np
import pandas as pd

def checkFlags_current(var):
    rows = len(var)
    # 初始化一维数组，避免后续展平操作
    flags = np.zeros(rows, dtype=int)

    # 先批量处理空值和0值，效率更高
    flags[pd.isnull(var)] = 3
    flags[var == 0] = 2

    # 循环范围限制到rows-2，避免访问i+1越界
    for i in range(rows - 1):
        # 修正运算符优先级，用括号包裹比较表达式，使用逻辑and
        if (var.iloc[i] > 0) and (abs(var.iloc[i+1] - var.iloc[i]) > 30):
            flags[i] = 1

    return flags.tolist()

# 调用方式不变
tmd_flags['Load Current(A)'] = checkFlags_current(tmd.iloc[:, 10])

方案2：使用Pandas向量化操作（推荐） #

Pandas的向量化操作比循环效率更高，且避免索引问题：

import pandas as pd

def checkFlags_current(var):
    # 初始化flags为0
    flags = pd.Series(0, index=var.index)
    
    # 批量设置各条件对应的flag值
    flags[var.isna()] = 3
    flags[var == 0] = 2
    # 计算差值条件：当前值>0，且下一个值与当前值的差的绝对值>30
    diff_condition = (var > 0) & (var.diff().abs().shift(-1) > 30)
    flags[diff_condition] = 1
    
    return flags.tolist()

# 调用方式不变
tmd_flags['Load Current(A)'] = checkFlags_current(tmd.iloc[:, 10])

内容的提问来源于stack exchange，提问作者Shraddha Jadhav