ncol Error in TS Model Forecast

[afabas]

Hi,

I am a novice user of the tool and have been handed a workflow from a previous colleague that was operational. I am getting an error at the TS Model Factory Module.  "Tool #64; Error in ncol(previous.output.table)".

I am trying to create an ETS time series model forecast for multiple SKUs based on a 3-year monthly sales history.  

I am inputting the Date, QTY, and SKU in. The QTY is set as a Fixed Decimal, the SKU is set as a V_String, the Date is set as Date. It seems to fulfill the input requirements.  A screenshot of sample input date is attached. 

The only output is the forecast parameters for a single SKU and not the full set.

I am at a loss as to what I need to look at next.

Thanks!

[CailinS]

Hello @afabas I hope I can provide a few next steps that may get you to resolution. They will come in the form of some questions and reviewing a troubleshooting blog I wrote a few (many) years ago that I think may help here.

1. Without seeing the actual configuration of the tool it's hard to know exactly what may be going wrong, so if you can provide your workflow or a screenshot of the config that may help
2. Here is a link to the Help for this tool which may provide some insight
   • https://help.alteryx.com/current/en/designer/tools/time-series/ts-model-factory-tool.html 
   • I have seen field type issues cause unexpected errors so one first guess is that...I can't tell what field type New PCN is, which I assume is the SKU, or Grouping field, and per the Help it must be string "The TS Model Factory tool requires a Designer data stream with at least two fields: the group name, which must be a String, VString, W_String, or V_WString, and the target field, which must be numeric. "
3. If the field type isn't the issue, here is the article that may clarify some of the next steps: https://community.alteryx.com/t5/Data-Science/Troubleshooting-the-Predictive-Tools/ba-p/8593 specifically, read down to the "Now for the fun stuff" section.
   • I'm looking to understand what messages the tool (macro) may be sharing beyond the one that bubbled up with the default message settings which don't pass through all messages.
4. Read through that article for other common data issues that may cause unexpected errors, and hopefully with the Macro Messages turned 'on' then you can get more info as to what is happening in the R process ahead of the error message that you reported above.

I'll keep an eye on your responses and any new information you can share about what you've tried!