Merging Datetime column and Int column as hours replacing 00 - ValueError: to assemble mappings requires at least that [year, month, day]

Question

我有一个包含两列的数据框：

具有以下格式的 datetime 列 %Y-%m-%d
int 列

它们代表一天和一个小时。我想合并，因为那时我想给 prophet 所以我需要转换成这种格式：

YYYY-MM-DD HH:MM:SS

我这样试过：

example={"date":["2018-04-18","2018-04-18","2018-04-18"],"alert_h":[4,17,23]}

df=pd.DataFrame(example)

df['date']=pd.to_datetime(df['date']) #I am converting str to datetime just for creating the example, I have already in the original df a datetime format

df.assign(date_h=pd.to_datetime(df[['date','alert_h']], format='%Y-%m-%d %h'))

我收到以下错误消息。

ValueError: to assemble mappings requires at least that [year, month, day] be specified: [day,month,year] is missing

我阅读了以下问题：

Value Error at least that year month day must be specified
How to convert columns into one date time column in pandas
Convert Pandas Column to DateTime

如果这是重复的，请随时关闭这个问题我会研究任何其他问题 link 因为我找不到解决我的问题的具体问题。

Answer 1

第一种方法：

newdf=df['date'].str.split('-',expand=True).assign(hour=df['alert_h'])
newdf=newdf.astype(str)
newdf.columns=['year','month','day','hour']
df['date_h']=pd.to_datetime(newdf,format='%Y-%m-%d %h:%m:%s')

第二种方法：

example={"date":["2018-04-18","2018-04-18","2018-04-18"],"alert_h":[4,17,23]}
df=pd.DataFrame(example)
date_h=df['date']+' '+df['alert_h'].astype(str)+':00:00'
df['date']=pd.to_datetime(df['date'])
df['date_h']=pd.to_datetime(date_h)

编辑（第三种方法）：

方法@MrFuppes（效率更高）：

df['date_h']=pd.to_datetime(df['date']) + pd.to_timedelta(df['alert_h'], unit='h')

Merging Datetime column and Int column as hours replacing 00 - ValueError: to assemble mappings requires at least that [year, month, day]

Merging Datetime column and Int column as hours replacing 00 - ValueError: to assemble mappings requires at least that [year, month, day]

python

pandas

datetime

facebook-prophet