Best Practice for Handling Time Series Gaps in KNIME?

Hi KNIME community,

I’m working on a time series dataset (hourly sensor data) and noticed there are irregular gaps in the timestamps. What’s the best node or strategy to fill in missing time steps in KNIME before feeding it into a forecasting model like Prophet or ARIMA?

I’m currently using “Missing Value” but not sure if that’s the right way to re-index time or interpolate missing rows. Would love to know how others handle this.

Thanks in advance!

Jhonn Mick

use the new labs nodes for timeseries.
given hourly data i wouldnt recommend writing the last value forward but instead using a mean

but without any further context what sensor it is and what type of data is being recorded, assumptions are barely useful

2 Likes

Thanks @fe145f9fb2a1f6b, I’ll check out the new Labs nodes. Appreciate the tip on using mean instead of forward-fill!

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.