Usage of variant attributes during import

The use of the variant attributes during merge and attribute mapping is not forbidden but strongly discouraged. The values of variant process attributes are overwritten by the variant calculation, which is only happening after all other calculations have already been processed. So neither one can use the results of the variant calculation nor the fully processed EPCs contain the values set outside the variant calculation.

There is a special danger in filling the dimension attributes by calculation rules or through mapping: When you do that while the variant dimension is not registered at the process type of the EPC, these values will not be overwritten. When you then register the dimension, re-initialize the analysis server, and do not recalculate the EPC, it will contain invalid dimension IDs, for example, a variant corresponding to the values would be shown on the GUI, but it would not correspond to any real variant in the database. In order to rectify the situation, you can runppmimport with –keyindicator new. (Using the command line parameter –ps to specify a suitable query, you can restrict the recalculation to the EPCs of the process type.)

Detailed information on how to use runppmimport can be found in the PPM Operation Guide.