REGISTERMODEL Procedure
PROC REGISTERMODEL Statement
The PROC REGISTERMODEL statement invokes the procedure. Table 1 summarizes the options in this statement.
Table 1: PROC REGISTERMODEL Statement Options
| Option | Description |
|---|
|
ALGORITHM= | Specifies the algorithm that was used to train the model |
|
DATA= | Specifies the name of the input table that was used to train the model |
|
DESCRIPTION= | Specifies the description of the model |
|
FUNCTION= | Specifies the model function type |
|
MODELER= | Specified the name of the modeler |
|
NAME= | Specifies the name of the model |
|
OAUTHTOKEN= | Specifies the name of the authentication token for connecting to SAS Model Manager |
|
REPLACE | Specifies that you want to replace existing models |
|
SERVER= | Specifies the root URL for your SAS Viya server where SAS Model Manager is located |
You must specify the following options:
-
NAME='model-name'
specifies the name to register the model under in SAS Model Manager. The 'model-name' must be a quoted string.
-
OAUTHTOKEN='oAuthToken-name'
specifies the name of the authentication token for connecting to SAS Model Manager. The 'oAuthToken-name' must be a quoted string.
-
SERVER='server-address'
specifies the root URL for your SAS Viya server where SAS Model Manager is located. The 'server-address' must be a quoted string.
You can also specify the following options:
-
ALGORITHM=algorithm-key-value | 'algorithm-name'
-
specifies the algorithm that was used to train the model. This can be specified as a key value from Table 2 or as a quoted string. When you specify the algorithm by using one of the listed key values, SAS Model Manager displays the algorithm name that is associated with the key value text in the language of the browser locale. When you specify it as a quoted string that does not match any of the listed key values, the algorithm name is treated as a user-defined value, and the value is not mapped to the display name.
Table 2: PROC REGISTERMODEL algorithm-key-values
|
algorithm-key-value | English Display Name |
|---|
| BINNING | Binning |
| BNET | Bayesian network |
| BOOLRULE | Boolean rules |
| ENSEMBLE | Ensemble |
| FACTMAC | Factorization machine |
| FOREST | Forest |
| GAM | Generalized additive model |
| GLM | Generalized linear model |
| GRADBOOST | Gradient boosting |
| IMPUTE | Imputation |
| LINEARREG | Linear regression |
| LOGISTIC | Logistic regression |
| MWPCA | Moving window principal component analysis |
| NEURAL | Neural networks |
| NPLR | Nonparametric logistic regression |
| OTHER | Other |
| PCA | Principal component analysis |
| PLSREG | Partial least squares regression |
| QUANTREG | Quantile regression |
| REPLACE | Replacement |
| RPCA | Robust principal component analysis |
| RULEBASED | Rule-based |
| SVD | SVD |
| SVDD | Support vector data description |
| SVM | Support vector machine |
| TEXTTOPIC | Text topic (discovery) |
| TREE | Decision tree |
-
DATA=libref.data-table
-
names the input data table for PROC REGISTERMODEL to use. The default is the most recently created data table. libref.data-table is a two-level name, where
- libref
refers to a collection of information that is defined in the LIBNAME statement and includes the library, which includes a path to the data. For more information about libref, see the section Using SAS Viya Workbench.
- data-table
specifies the name of the input data table.
For a DATA step model, you must provide an input data table containing all the input variables that the model uses. You also must provide an input data table if you specify the ASSESSMENT statement.
-
DESCRIPTION='model-description'
specifies the description of the model. The 'model-description' must be a quoted string.
-
FUNCTION=function-key-value
-
specifies the model function type. You can specify one of the key values that are listed in Table 3, which also lists the corresponding display names. SAS Model Manager displays the function name that is associated with the key value text in the language of the browser locale.
By default, FUNCTION=ANALYTICAL.
Table 3: PROC REGISTERMODEL function-key-values
|
function-key-value | English Display Name |
|---|
| ANALYTICAL | Analytical |
| CLASSIFICATION | Classification |
| CLUSTERING | Clustering |
| IMAGE_CLASSIFICATION | Computer vision – image classification |
| IMAGE_SEGMENTATION | Computer vision – image segmentation |
| KEYPOINTS_DETECTION | Computer vision – keypoints detection |
| OBJECT_DETECTION | Computer vision – object detection |
| PREDICTION | Prediction |
| SEGMENTATION | Segmentation |
| SINGLE_SERIES_FORECASTING | Single-series forecasting |
| TEXT_CATEGORIZATION | Text categories |
| TEXT_CONCEPTS | Text concepts |
| TEXT_SENTIMENT | Text sentiment |
| TEXT_TOPICS | Text topics |
| TRANSFORMATION | Transformation |
Only the classification and prediction functions are supported for assessment.
-
MODELER='modeler'
-
specifies the name of the modeler. The 'modeler' must be a quoted string.
By default, this option is set to the value of the SYSUSERID automatic macro variable. For more information about the SYSUSERID automatic macro variable, see SAS Macro Language: Reference.
-
REPLACE
specifies that you want to replace an existing model with a new one if the new model would have the same model name in the same project.
Last updated: September 23, 2026