With model level data from the AUS2200 files have two time axis, for example
each file at 1200 hours have variable defined at 1200 hours but also variable defined at 1300 hour.
This causes issue when only timestamp is used to work out time range. So adding check in get_timedim to returned if multiple times are in axis
and then using check_in_range instead of check_timestamp if multiple_times is True.
This seems to work but we also have to now adapt the set(time) after calculation to make sure we're not cutting off day after at 00 hour where this occurs.
So in cli.py
dsin = dsin.sel({time_dim: slice(ctx.obj['tstart'], ctx.obj['tend'])})
and later on
out_var = out_var.sel({time_dim: slice(ctx.obj['tstart'], ctx.obj['tend'])})
With model level data from the AUS2200 files have two time axis, for example
each file at 1200 hours have variable defined at 1200 hours but also variable defined at 1300 hour. This causes issue when only timestamp is used to work out time range. So adding check in get_timedim to returned if multiple times are in axis and then using check_in_range instead of check_timestamp if multiple_times is True. This seems to work but we also have to now adapt the set(time) after calculation to make sure we're not cutting off day after at 00 hour where this occurs. So in cli.py dsin = dsin.sel({time_dim: slice(ctx.obj['tstart'], ctx.obj['tend'])}) and later on out_var = out_var.sel({time_dim: slice(ctx.obj['tstart'], ctx.obj['tend'])})