Question

我正在尝试从Pandas的庞大数据集中删除一些行。我决定使用iterrows（）函数搜索要删除的索引（因为我知道在迭代时删除是个坏主意）。现在看起来像这样：

list_to_delete = []
rows_to_delete = {}
for index, row in train.iterrows():
    if <some conditions>:
        list_to_delete.append(int(index))
        rows_to_delete[int(index)] = row
train = train.drop([train.index[i] for i in list_to_delete])

这给了我这样的错误：

Traceback (most recent call last):
  File "C:/Users/patka/PycharmProjects/PARSER/getStatistics.py", line 115, in <module>
    train = train.drop([train.index[i] for i in list_to_delete])
  File "C:/Users/patka/PycharmProjects/PARSER/getStatistics.py", line 115, in <listcomp>
    train = train.drop([train.index[i] for i in list_to_delete])
  File "C:\Users\patka\PycharmProjects\PARSER\venv\lib\site-packages\pandas\core\indexes\base.py", line 3958, in __getitem__
    return getitem(key)
IndexError: index 25378 is out of bounds for axis 0 with size 25378

怎么可能？

在此之前，我创建了该数据集的副本，并尝试在遍历原始副本（使用inplace = True）的同时从该副本中删除选定的行。不幸的是，有一个错误消息说NoneType对象没有属性'drop'。

非常感谢您的帮助。我的示例行如下所示：

resolution Done 
priority Major 
created 2000-07-04T13:13:52.000+0200 
status Resolved 
Team XBee 
changelog {'Team" : {'from':...

从条件中删除Python Dataframe中的行

0 个答案: