The Manual Column Specification Processor is an auxiliary processor that gives the user control over the column information that is currently present at the client Type Inference (TI). The user can remove columns that are already inside the TI and add new columns with a certain scale and representation type (double, integer, numeric, string and datetime).
Note, that this processor does not trigger the generation of new columns on server side but is only used to aid client TI. Subsequent processors will fail when selected columns are generated virtually by this processor but are not present in their physical input dataset.
The processor can operate on any dataset.
The processor output matches the input dataset but makes it possible to directly refer to the columns that were created in a previous processor without having to save the resulting dataset with the new columns first.
The short video below shows an example of type inference problems that can be solved using the Manual Column Specification processor.
The example above shows how a query inserted in the Query Processor is used to aggregate the dataset and generate a new column called "mean_calc".
If we want to use the Exclude Columns Processor in the next step, this processor now still shows us the columns "A" and "B" of the original dataset and not the columns "A" and "mean_calc".
We can change this using the Manual Column Specification processor by deciding to exclude the column "B" from the type inference and adding the columns "mean_calc" to it.