You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
CSVMixin should behave like PIMixin, but does not.
The relevant part in CSVMixin's set_timeseries:
ifisinstance(timeseries, Timeseries):
# TODO: add better check on timeseries.times?ifcheck_consistency:
ifnotnp.array_equal(self.times(), timeseries.times):
raiseException(
'CSV: Trying to set/append timeseries {} with different times ''(in seconds) than the imported timeseries. Please make sure the '# ...
It should not check against self.times(), but against self.__timeseries_times_sec. What happens now is that doing a get_timeseries will return an object with len(ts.times) != len(ts.values):
In GitLab by @vreeken on Sep 7, 2019, 13:32
Calling
set_timeseries
with aTimeseries
objectCSVMixin should behave like PIMixin, but does not.
The relevant part in CSVMixin's
set_timeseries
:It should not check against
self.times()
, but againstself.__timeseries_times_sec
. What happens now is that doing aget_timeseries
will return an object withlen(ts.times) != len(ts.values)
:In PIMixin, there's the following check:
All missing values are filled with NaNs (before and after).
Calling
set_timeseries
with annp.ndarray
object:Now for the case where we call
set_timeseries
with an array.PIMixin has:
This is not sufficient. We should also check that
self.__timeseries_import_times
is a superset ofself.times()
.Same goes for CSVMixin:
Again should check that it is a superset.
The text was updated successfully, but these errors were encountered: