0

I´m using Analysis Service to make some data mining work. I already made the predictions with Sql Server Data Tools. Now i wanted to make the same prediction with dmx.

I´m using the next dmx code:

  CREATE MINING MODEL [Cassandra]
(
   [CompanyNK] TEXT KEY,
   [Date] Date KEY TIME,
   [Total Value Mes] Double Continuous Predict,
   [New Items Month] Long Continuous,
   [Number Branches] Long Continuous,
   [Number Clients] Long Continuous,
   [Number Salesman] Long Continuous

)
USING Microsoft_Time_Series(AUTO_DETECT_PERIODICITY = 0.8, FORECAST_METHOD = 'MIXED')
WITH DRILLTHROUGH

But i get the following error:

Error (Data mining): Usage of all non-key columns in the 'Cassandra' mining model must be set to PredictOnly.

This should work, because in SSDT i only have one variable set to "predict" and i have none set to "predict only". Some help would be awesome.

Thanks in advance. Helder Borges

hoss
  • 2,430
  • 1
  • 27
  • 42
Helder
  • 3
  • 3

1 Answers1

0

Ok... I don't know what the problem was. But shutting-down the computer for all the weekend appeared to solve the problem. I didn't do anything despite of turning the computer on and compile, and it worked.

Best Regards, Helder Borges

Updtae: Found that the problem was due to MS-SSAS version. You are only able to use predictions for prediction in the enterprise (or BI) edition. And i was using the standard.

Helder
  • 3
  • 3