Using SPSS Modeler, I'm importing a SAS file, adding additional fields, and exporting the final result to a text file. I'm having difficulties with SPSS changing text fields from the imported SAS file to numerical fields.
Several fields from the SAS file are text files but contain only numbers. SPSS automatically converts these to Continuous measurement, and then adds ".00" at the end of the value. This causes problems in my downstream process. I added a Type node and manually set the problematic fields to Categorical measurement, but Modeler keeps adding the ".00" anyway (I assume it changes them back to Numeric after the Type node).
How can I preserve the original field type (numerical value stored as text) and prevent these being converted to numeric fields?
I'm running Modeler 18.0 on a Windows 7 machine.
To change the storage type of the field, you need to use a Filler node. Select the affected fields in the "Fill in fields", set the Replace option to "Always" and the "Replace with" to
This will convert the selected fields to string fields without any decimals.
Thank you. Your suggestion works.
I wish Modeler was more consistent when importing data. The Var. File node (for text files) allows the user to override the Storage type of each field (in the "Data" tab of the input node). However, this is not available in most other input nodes, requiring the workaround specified above.
JDBC Export 1 Answer