reading forecasts for split file variables
Hi,<br>
I have an input file with two variables: the first one (V1) is a name of a metric, and the second one (V2) if a measured value for V1 metric. Each metric has many values. I use SPLIT FILE on V1 variable and then calculate forecasts using V2 as dependent variable (f.e. using Forecasting's Expert Modeler).<br>
In viewer I receive separate table for each SPLIT FILE variable. But using <br>
forecast = spssaux.getValuesFromXmlWorkspace(dsc_table, 'Forecast', rowCategory="Forecast", cellAttrib="number")<br>
I receive one big list with forecasts for each SPLIT FILE varaible (all forecasts are in one list).<br>
Is there possibility to read forecasts for one specific SPLIT FILE (V1) variable? How can I detect number of predicted forecasts for each V1 variable or check variables for which forecasts were calculated?<br>
Regards,
Lukasz
Please post an spv file containing a table of the type you are trying to retrieve from. It might be that you will have to use the lower level spss.EvaluateXPath api, but let's first torture the getValues api.
Thanks Jon.<br>
I attached the file I use.
I'm not following what you are doing.<br>
You can't split by v1 in your data, because it is too long. Are you autorecoding it first?<br>
And are you using SPLIT FILE SEPARATE?<br>
Which output table(s) are you actually retrieving, and how are you generating the forecasts? You can't use Apply TS models when split files is on.<br>
If you are retrieving the parameter table to do the forecasting yourself, are you checking whether the EM chose exponential smoothing or ARIMA?<br>
The most helpful clarification would be for you to post an spv with the actual output table(s) you need to retrieve.
Regards,
Jon