SARIMA Temperature Forecasting

This workflow demonstrates how the SARIMA components can be used to generate forecasts. In this case for hourly temperature data.

This is a companion discussion topic for the original entry at


I wanted to forecast using automatic SARIMA but I am getting this notification

ERROR Breakpoint 7:56:0:84:0:346:347 Execute failed: Seasonal component settings incompatable, verify S and at least one other term are non-zero, or all are zero.

How can i solve this problem.


Hi @gizemyigi and welcome to the forum.

I am able to run this workflow without any changes as downloaded from the Hub. Are you making changes to the settings, or trying to use a different input file?

If so, maybe it would best to post a new thread on the main Analytics Platform with more details (and if possible, uploading your workflow in progress):

1 Like

Hi @ScottF

Why the workflow isn’t loading?
The forecast process freezes in the loading phase.


It looks like the workflow is loading OK, but is stuck on execution of the Auto-SARIMA component.

How long did you let the component run? It can take quite a while depending on the input parameters. Do you see any message in the console or the log?


About 15 minutes

ready charge however it took a long time. can i influence my pc?

Check here for how you can allocate more RAM to KNIME:

You may also just need a faster CPU, depending on your hardware.


@ScottF What a shame to bother one last time, I want to know if it is possible for the forecast to give me an interval instead of a single value

Something like this?

One of our other data scientists has made the preliminary updates to the component, but we need to do a bit of verification and testing before it gets pushed to the public KNIME Hub. So watch the Auto-SARIMA component for a future update!

(And thanks for the feedback!)


Also it just occurred to me that the main reason this component took so long to run on your machine was the Conda Environment Propagation node embedded inside, which is creating a new Python environment with the packages needed to run.

The good news is this is a one time operation, and subsequent executions of the component will use the already created environment. So future runs should be noticeably faster.


Thanks for all, I really appreciate your help. I can’t wait for it.

Hi @Jalvear -

The component has been updated to include the lower and upper bounds. Hope this helps!