Whether the program considers missing time ？

[mavarick]

Is your feature request related to a current problem? Please describe. for example, time series with specified time window by day, but some days are missing.

Describe proposed solution one way is checking time continuity before training/predicting, and fill empty dates with some values.

I want to know if there are some advanced or automatic ways to tackle this problem.

[ichsan2895]

I think you can use this way :

series = TimeSeries.from_dataframe(DF, 'your date column name', 'your target column name', fillna_value=a number)

[hrzn]

Another way is to use missing value fillers, for instance:

from darts.utils.missing_values import fill_missing_values
my_series = fill_missing_values(my_series)

which will interpolate the missing values.