Python Pandas 条件未能准确识别行

Question

这是来自 jupyter notebook 的输入和输出。我需要帮助来确定我无法准确 select 并在 'went_out' 列中设置数据的原因。

两个带红色下划线的单元格都应该显示其所在行的日期时间列中的数据，但只有一个单元格准确显示。事实证明，许多符合我条件的行没有得到 selected 和设置。

这是我使用的代码示例：

# your answer here
df.loc[(df['reading_type'] == 'motion') & (df['value'] == 255), 'event'] = 'motion on'
df.loc[(df['reading_type'] == 'motion') & (df['value'] == 0), 'event'] = 'motion off'

df2 = df.loc[(df['reading_type'] == 'door') | (df['event'] == 'motion on')].copy()
df2.loc[(df['event'] == 'door close') & (df['event'].shift(-1) == 'door open'), 'went_out'] = df2['datetime']
df2

这里是 jupyter notebook 文件和 csv 文件的链接：

Jupyter 笔记本： https://drive.google.com/file/d/15f6NQrM4UoAZlzRhK35TOKyhPJnmWWdU/view?usp=sharing
CSV 文件： https://drive.google.com/file/d/1hZudSVbT91ESj2qkzrJ--CbVdrzVCmce/view?usp=sharing

Answer 1

据我了解，您正试图写下关门的日期和时间。这可能是您想要的解决方案的一部分。您可以只使用门关闭条件来索引 'went_out' 列，而不是寻找门打开然后关闭的条件。

df.loc[(df['reading_type'] == 'door') & (df['value'] == 255), 'event'] = 'door on'
df.loc[(df['reading_type'] == 'door') & (df['value'] == 0), 'event'] = 'door off'

df2 = df[df['reading_type'] == 'door'].copy()
# The line below is modified
df2.loc[df2['event'] == 'door off', 'went_out'] = df2[df2['event'] == 'door off']['datetime']
print(df2)

输出如下：

    id  datetime    device  location    reading_type    value   event   went_out
284 284 2018-01-01 07:57:56 Door    door    door    255.0   door on NaN
285 285 2018-01-01 07:58:12 Door    door    door    0.0 door off    2018-01-01 07:58:12
294 294 2018-01-01 08:29:25 Door    door    door    255.0   door on NaN
295 295 2018-01-01 08:29:38 Door    door    door    0.0 door off    2018-01-01 08:29:38
357 357 2018-01-01 09:16:38 Door    door    door    255.0   door on NaN
361 361 2018-01-01 09:17:40 Door    door    door    0.0 door off    2018-01-01 09:17:40

希望这对您有所帮助。

编辑
开门后关门时获取日期和时间的条件

df2.loc[((df2['event'].shift(-1) == 'door on') & (df2['event']=='door off') ), 'went_out'] = df2[df2['event']=='door off']['datetime']

print(df2[df2['event'] == 'door off'])

    id  datetime    device  location    reading_type    value   event   went_out
285 285 2018-01-01 07:58:12 Door    door    door    0.0 door off    2018-01-01 07:58:12
295 295 2018-01-01 08:29:38 Door    door    door    0.0 door off    NaN
361 361 2018-01-01 09:17:40 Door    door    door    0.0 door off    2018-01-01 09:17:40
509 509 2018-01-01 15:50:46 Door    door    door    0.0 door off    2018-01-01 15:50:46

如果这能解决您的问题，请告诉我。

Python Pandas 条件未能准确识别行

Python Pandas condition fails to identify rows accurately

python

pandas

jupyter

data-science

jupyter-notebook