"The results of the current transformation cannot be shown in this type of plot"

[Irene]

I try to replicate the example 6.2 in Montgomery, Jennings and Kulahci's textbook. The dynamic regression model is: the error AR(1), transfer function w/1-delta1B-delta2B^2 and lag=3. After specifying

Y = P=( 1 )  NOINT  + x : LAG=3 DEN=( 1 2 )

in the SubsetARIMA model.

The results cannot be shown. It reads "The results of the current transformation cannot be shown in this type of plot".

Could anyone help? 

[udo_sas]

Hello -

After creating this model using the Subset ARIMA dialog - do you see any messages in the LOG file?

The reason for failing to show plot is usually data dependent, so you might be better of contacting Technical Support.

Thanks,

Udo

[Irene]

Thanks for the timely reply. I have attached here the log file. So can you see where the problem is?

Besides, may I ask how to use FS do the prewhitening? Or, in general, is there a quick way that FS can help identify the specifications of dynamic regression model?

[udo_sas]

Hello -

Yes, there are ERROR: messages at line 1861 (there may be more). Please contact Technical Support for additional help.

With regards to your second question: you will find details of how SAS Forecast Studio creates dynamic regression models automatically in SAS Forecast Server Procedures 14.1: User's Guide (check out details section of HPFDIAGNOSE procedure, covers prewhitening too).

Thanks,

Udo

[alexchien]

I am not sure if the ID statement is specified correctly

id time interval=DTMINUTE format=TIMEAMPM11. acc=total notsorted horizonstart=&HPF_HORIZON_START start=&HPF_START;

Are you trying to forecast a series in the minutes interval? The input data contains 100 observations but the output appears to be 8 millions+ observations.

ERROR: Unable to allocate requested memory for time series meta-model _LEAF_3 for variable y.

WARNING: No model was chosen from the selection list for variable y. Now attempting to forecast using exponential smoothing.

WARNING: Unable to forecast the variable y. Forecast values set to missing.

NOTE: Number of series processed: 1

NOTE: Number of forecasts requested: 1

NOTE: Number of forecasts failed: 1

NOTE: There were 100 observations read from the data set _LEAF.DATA.

NOTE: The data set WORK.OUT has 8726411 observations and 2 variables.

NOTE: The data set WORK.OUTFOR has 8726411 observations and 8 variables.

[Irene]

I am using SAS Forecast Studio 13.1 for Desktop. So I cannot code and use any procedures? And, I cannot automatically get the dynamic regression model specifications and forecasting, right?

[udo_sas]

Hello -

"I am using SAS Forecast Studio 13.1 for Desktop. So I cannot code and use any procedures?"

Correct.

"" And, I cannot automatically get the dynamic regression model specifications and forecasting, right?

You should see the automatic dynamic regression model specifications in model view.

Thanks,

Udo