select a specific time with and create netcdf with xarray

[FRANCOIS100]

I hereby ask for your help in python. In fact, I have the NetCDF data from IMERG / GPM (Global precipitation measurement) which has 30 min of temporal resolution and I would like to calculate the sum or aggregations event rainfall on particular dates where I know the start and the end of each date on which I want to aggregate.
I have already tried to do this without success with the resample function for each start timestamp to end timestamp and save it in netcdf file.
I have attached my sript and the .csv file which has the start and end date for each aggregation.

data1 = xr.open_dataset('/home/fx/GPM/latlon_mokabi_2016_2019_GPM.nc')

time_start_and_end =  np.loadtxt("/home/fx/thesis_analyse/analyse_11_01_2021/select/makabi_timestamp")

lat = data1.lat
lon = data1.lon

nlin, ncol = np.shape(df_clus1)
# print(data1)
event = []
for i in range(0,nlin,1):
    a = int( time_start_and_end[i, 0])
    b = int( time_start_and_end[i, 1])
    # print(a)
    # print(b)
    event_select = data1['precipitationCal'][a:b][:,:].sum('time')
    event.append(event_select)

print(event)

ds = xr.Dataset({'precipitationCal': (['time', 'lat', 'lon'], event)},
               coords={'lon': (['lon'], lon),
                         'lat': (['lat'], lat),
                         'time': time},)

[keewis]

we don't (yet?) support indexing by a sequence of slices, so if your data is irregular this might be difficult. If it is regular, rolling or resample might be enough. Can you show the repr of data1 and some examples for time_start / time_end?

[FRANCOIS100]

My time is irregular so rolling can't be working here is each start and end dates/times I would like to calculate the sum or aggregations event rainfall from each start and end dates below

Datetime_start | Datetime_end
2012-03-05 04:00:00 | 2012-03-05 07:30:00
2012-05-13 21:30:00 | 2012-05-14 06:30:00
2012-05-19 01:30:00 | 2012-05-19 02:30:00
2012-05-20 02:30:00 | 2012-05-20 04:30:00
2012-05-21 19:30:00 | 2012-05-21 20:30:00
2012-05-31 20:00:00 | 2012-05-31 23:30:00
2012-12-11 17:30:00 | 2012-12-11 18:30:00
2012-12-26 01:30:00 | 2012-12-26 04:00:00
2013-06-01 20:00:00 | 2013-06-01 21:30:00
2013-01-16 01:00:00 | 2013-01-16 04:30:00
2013-01-18 07:00:00 | 2013-01-18 10:00:00
2013-01-25 21:30:00 | 2013-01-25 22:30:00
2013-01-26 18:00:00 | 2013-01-26 21:30:00
2013-01-02 19:00:00 | 2013-01-02 23:30:00
2013-02-02 15:00:00 | 2013-02-02 16:30:00
2013-02-17 18:30:00 | 2013-02-17 20:30:00
2013-02-18 02:30:00 | 2013-02-18 06:30:00
2013-03-14 17:30:00 | 2013-03-14 22:00:00

[mathause]

groupby_bins could work for this or a loop.