使用 eval 时出错： "unknown type timedelta64[ns]" -- Python

Question

我正在尝试使用"eval"；使用以下代码：

delay = (
dataset
.query("INCIDENT_TYPE_DESC == '111 - Building fire'")
.eval("ARRIVAL_DATE_TIME - INCIDENT_DATE_TIME")
.divide(pd.Timedelta("1m"))
)

但我得到以下错误：

ValueError: unknown type timedelta64[ns]

请在下面找到数据集信息：

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1736391 entries, 0 to 1736390
Data columns (total 24 columns):
#   Column                        Dtype         
---  ------                        -----         
0   IM_INCIDENT_KEY               object        
1   FIRE_BOX                      object        
2   INCIDENT_TYPE_DESC            object        
3   INCIDENT_DATE_TIME            datetime64[ns]
4   ARRIVAL_DATE_TIME             datetime64[ns]
5   UNITS_ONSCENE                 float64       
6   LAST_UNIT_CLEARED_DATE_TIME   object        
7   HIGHEST_LEVEL_DESC            object        
8   TOTAL_INCIDENT_DURATION       float64       
9   ACTION_TAKEN1_DESC            object        
10  ACTION_TAKEN2_DESC            object        
11  ACTION_TAKEN3_DESC            object        
12  PROPERTY_USE_DESC             object        
13  STREET_HIGHWAY                object        
14  ZIP_CODE                      object        
15  BOROUGH_DESC                  object        
16  FLOOR                         object        
17  CO_DETECTOR_PRESENT_DESC      object        
18  FIRE_ORIGIN_BELOW_GRADE_FLAG  float64       
19  STORY_FIRE_ORIGIN_COUNT       float64       
20  FIRE_SPREAD_DESC              object        
21  DETECTOR_PRESENCE_DESC        object        
22  AES_PRESENCE_DESC             object        
23  STANDPIPE_SYS_PRESENT_FLAG    float64       
dtypes: datetime64[ns](2), float64(5), object(17)
memory usage: 317.9+ MB

数据集描述/信息上方。我正在尝试从INCIDENT_DATE_TIME和ARRIVAL_DATE_TIME 中减去数据

Answer 1

我在尝试使用eval时遇到了同样的错误。也许你可以试试dataset['ARRIVAL_DATE_TIME'] - dataset['INCIDENT_DATE_TIME']而不是