The CCDM Procedure

SEVERITYMODEL Statement

SEVERITYMODEL severity-model-list </ definition-option>;

The SEVERITYMODEL statement specifies one or more severity distribution models that you want to use in simulating a compound distribution sample. The severity-model-list is a space-separated list of names of severity models that you would use with PROC SEVSELECT’s DIST statement. The SEVERITYEST= data table or the SEVERITYSTORE= item store must contain all the severity models in the list. If you specify the SEVERITYEST= data table and you specify a name that does not appear in the _MODEL_ column of the SEVERITYEST= data table, then that name is ignored. Similarly, if you specify the SEVERITYSTORE= item store and a severity model by that name does not appear in the item store, then that name is ignored.

You can specify the following definition-option in the SEVERITYMODEL statement:

DEFINITIONSTABLE="CAS-table-name" <(CASLIB="caslib")> SEVDEF="CAS-table-name" <(CASLIB="caslib")>

specifies a data table on the CAS server that contains the severity distribution function definitions. If you do not specify the CASLIB= option, then PROC CCDM assumes that the table is in the caslib that is active in the current CAS session. To specify a different caslib, use the CASLIB= option.

If you do not specify the DEFINITIONSTABLE= option, then PROC CCDM loads the severity function definitions from the library Sashelp.Svrtdist and the libraries that you specify in the CMPLIB= SAS system option and creates a temporary CAS table named cdm_svrtdist in the current CAS session. This process requires some time and effort, which you can avoid by specifying the DEFINITIONSTABLE= option. You can use the CAS table that the SEVSELECT procedure creates when it fits the severity models. For more information, see the OUTFUNCDEF= option in the DIST statement of the SEVSELECT procedure.

If you specify more than one SEVERITYMODEL statement, only the first one is used.

If you do not specify a SEVERITYMODEL statement, then this is equivalent to specifying all the severity models that appear in the SEVERITYEST= data table or the SEVERITYSTORE= item store.

A compound distribution sample is generated for each severity model by compounding that severity model with the frequency model that you specify in the COUNTSTORE= item store or the COUNTMODEL statement, or the external frequency model that is encoded by the COUNT= variable that you specify in the EXTERNALCOUNTS statement.

Last updated: April 15, 2021