Project Properties
About Project Properties
Project properties contain
the project metadata. Project metadata includes information such as
the name of the project, the project owner, the project identifier,
the name and path of the repository, and of the tables and variables
that are used by project processes.
Project properties are
organized into the following types:
-
General Properties
-
Specific Properties
-
System Properties
-
User-Defined Properties
General Properties
General
Properties are system-defined properties that you cannot
modify, with the following exceptions: folder description, operation
status, and lock status.
|
Property Name
|
Description
|
|---|---|
|
Model function
|
Specifies the type of
output that your predictive model project generates. The Model
function property that you specify affects the model
templates that are provided when you are ready to import models into
one of your project's version folders. After it has been declared,
the Model function property for a project
cannot be changed. Ensure that the types of models that you are going
to use in the project fit within the selected model function type.
For more information about the types of model functions, see
Types of Model Functions.
|
|
Operation
status
|
Specifies the current
state of the project:
Under Development
indicates that the
project has started but a champion model is not yet in production.
Active
indicates that a champion
model for this project is in production.
Inactive
indicates that the
champion model is temporarily suspended from production.
Retired
indicates that the
champion model for this project is no longer in production.
To set the status, select
the Operation status.
|
|
Lock project
variables
|
Specifies that the project
metadata is locked and the project definition cannot be modified.
For more information, see
Lock or Unlock Project Variables.
|
|
Model Function
|
Description
|
Example
|
|---|---|---|
|
Analytical
|
Function for any model
that is not Prediction, Classification, or Segmentation.
|
None
|
|
Classification
|
Function for models
that have target variables that contain binary, categorical, or ordinal
values.
|
DEFAULT_RISK =
{Low, Med, High}
|
|
Prediction
|
Function for models
that have interval targets with continuous values.
|
The score output of
a prediction model could estimate the weight of a person. The output
of a model would be P_Weight.
|
|
Segmentation
|
Function for segmentation
or clustering models.
|
Clustering models
|
Specific Properties
Specific
Properties contain information about tables that are
used by the project as well as various input and output variables
and values that are used in scoring the models in test and production
environments. This data can be added or modified after you add variables.
For more information, see
Defining Project Input and Output Variables.
|
Property Name
|
Description
|
|---|---|
|
Default
test table
|
Specifies a default
SAS data set that can be used to create the New Dynamic Lift and Interval
Target Variable reports.
|
|
Default
scoring input table
|
Specifies a default
SAS data set that is used as the input data table for all scoring
tests within the project. If you specify a value for the Default
scoring input table property, the value is used as the
default input table in the Add a New Scoring Test window.
|
|
Default
scoring output table
|
Specifies a default
SAS data set that defines the variables to keep in the scoring results
table and the scoring test output table. If you specify a value of
the Default scoring output table property,
the value is used as the default output table in the Add
a New Scoring Test window.
|
|
Default
performance table
|
Specifies the default
performance table for all model performance monitoring tests within
a project.
|
|
Default
train table
|
The train table is optional
and is used only as information. However, when a value is specified
for a model's Default train table property,
the default train table is used to validate scoring functions or scoring
model files when a user publishes the associated project champion
model or challenger models to a database.
|
|
Champion
version
|
Specifies the version
that contains the champion model in a production environment.
|
|
Model function
|
Specifies the type of
output that your predictive model project generates. The Model
function property that you specify affects the model
templates that are provided when you are ready to import models into
one of your project's version folders. After it has been declared,
the Model function property for a project
cannot be changed. Ensure that the types of models that you use in
the project fit within the selected model function type.
|
|
Training
target variable
|
Specifies the name of
the target variable that was used to train the model.
|
|
Target event
value
|
Specifies the target
variable value that defines the desired target variable event.
|
|
Class target
values
|
For class, nominal,
ordinal, or interval targets, the set of possible outcome classes,
separated by commas. For example, binary class target values might
be
1, 0 or Yes, No.
Nominal class target values might be Low, Medium,
High. These values are for information only.
|
|
Class target
level
|
Specifies the class
target level of binary, nominal, ordinal, or interval.
|
|
Output event
probability variable
|
The output event probability
variable name, when the Model function property
is set to Classification or Analytical.
|
|
Output prediction
variable
|
The output prediction
variable name, when the Model function property
is set to Prediction or Analytical.
|
|
Output segmentation
variable
|
The output segmentation
variable name, when the Model function property
is set to Classification, Segmentation or Analytical.
|
Copyright © SAS Institute Inc. All rights reserved.