根据 Pandas 中的整数值将 DataFrame 行拆分为多行

Question

Is there way we can split a row into multiple rows based on an integer cap value? I have a dataframe as below

sys_df = pd.DataFrame([{'dateTime': '2020-11-12 17:45:00', 'timeTakenInSeconds': 650, 'id':'xyz'}])

Index	dateTime	timeTakenInSeconds	id
0	2020-11-12 17:45:00	650	xyz

I am trying to split the above row into 3 rows of previous 5 minute intervals like below.

Index	dateTime	timeTakenInSeconds	id
0	2020-11-12 17:45:00	300	xyz
1	2020-11-12 17:40:00	300	xyz
2	2020-11-12 17:35:00	50	xyz

Do we have any pandas builin utils to achieve this?

Answer 1

You can build your own method. A lead can be:

import pandas as pd
from datetime import datetime, timedelta
origin_data = {'dateTime': '2020-11-12 17:45:00', 'timeTakenInSeconds': 650, 'id':'xyz'}

def splitter(origin_data, interval=0):
    data=[]
    to_sec = interval*60
    current_time = datetime.fromisoformat(origin_data['dateTime'])
    for item in range((origin_data['timeTakenInSeconds']//to_sec)):
        data.append({'dateTime': str(current_time),
                     'timeTakenInSeconds': to_sec, 'id':'xyz'})
        current_time -= timedelta(seconds=to_sec)
    reminder = origin_data['timeTakenInSeconds'] - (origin_data['timeTakenInSeconds']//to_sec)*to_sec
    if reminder:
        data.append({'dateTime': str(current_time),
                     'timeTakenInSeconds': reminder, 'id': 'xyz'})
    return data


print(pd.DataFrame(splitter(origin_data, interval=5)))

Outputs:

              dateTime  timeTakenInSeconds   id
0  2020-11-12 17:45:00                 300  xyz
1  2020-11-12 17:40:00                 300  xyz
2  2020-11-12 17:35:00                  50  xyz

Note:

You can also use:

pd.date_range(end=datetime.fromisoformat(origin_data['dateTime']), periods=3, freq='5min')

To split the date as you wish.