如何將字符串數據以可變格式轉換爲datetime？

Question

我有了時間的列字符串格式如下一個熊貓數據幀：

id  playTime 

12  1d 13h 23m 
13  42d 22h 47m 
14  7d 16h 31m 
15  1h 26m 6s 
16   9d 58m 

我tried pd.to_datetime(df['playTime'], format='%j%H%M%S')但它表明time data does not match format '%j%H%M%S' (match). In the end, I'd like to convert the DF [「playTime內容」] values to seconds.

Answer 1

我覺得需要to_timedelta爲Timedelta s的total_seconds：

df['playTime'] = pd.to_timedelta(df['playTime']).dt.total_seconds().astype(int) 
print (df) 
    id playTime 
0 12 134580 
1 13 3710820 
2 14 664260 
3 15  5166 
4 16 781080 

Answer 2

你可以試試這個：

pd.to_datetime(df['playTime'], format='%j %H %M %S')