When you select the Time ID variable in step 4 of the SAS Forecast Studio New Project Creation wizard, you might receive the warning shown below:

By default, SAS Forecast Studio uses the first 10,000 observations to detect the time interval. Usually, a few distinct Time ID values are needed in order for the software to detect the interval. However, data sets with many repeated values of the Time ID variable sometimes require a larger sample in order to have enough distinct values. Increasing the number of observations might improve time-interval detection when you use data sets that contain many repeated values. Values in quantities of 100,000 or more can cause noticeably slower performance when you create a new project.

To configure the installed SAS Forecast Studio client so that it uses a different number of observations to detect the time interval:

1. Open the forecaststdo.ini file in a text editor. In a default Microsoft Windows installation, this file is located in the C:\Program Files\SAS\SASForecastStudio\12.1 directory.
2. In forecaststdo.ini, enter a new JavaArgs_n entry, where n is the number of the next argument in the list. Increment the argument number accordingly.

   Example:

   JavaArgs_14=-Dforecasting.interval_sample=50000
3. Relaunch the Forecast Studio Client.

To configure the SAS Forecast Studio Java Web Start client to use a number of observations to detect the time interval:

1. From SAS^® Management Console, open the SAS Forecast Server 12.1 properties, as shown below.

   

2. Add a semi-colon to the end of the Forecast Studio (Java Web Start) Other arguments section and append the parameters minus the -D, as shown in the detailed image of the Other arguments section below. Note: The -D is not required when you set the parameter for the Java Web Start client.

   

3. Click OK.
4. Restart the SAS Forecast Server 12.1 Web Application Server.
5. Verify the changes in the SAS Forecast Server Status window, as shown below:

   

Operating System and Release Information