The NEURAL Procedure
QUIT Statement
Stops the procedure.
Category Action Statement - affects the network or the data sets.
Options'>Options set in an action statement
affect only that statement.
QUIT;
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The NEURAL Procedure
RANOPTIONS_Statement__Specifies_distribution_of_random_initial_weights.__Category'>RANOPTIONS Statement
Specifies distribution of random initial weights.
Category Option Statement - does not
directly affect the network, but sets options for use in
subsequent action statements. The options persist until reset at a later stage in the
processing.
RANOPTIONS connection-list /
<RANDF=number>
<RANDIST=name>
<RANDOM=integer>
<RANLOC=number>
<RANSCALE=number>;
Required Arguments
Note: When a RANOPTIONS statement is executed, the specified options are stored in all the
connections listed before the slash. These options are used whenever an INITIAL or PRELIM statement
is executed. If you submit two RANOPTIONS statements for the same connection, the second statement
overrides all options in the first. In other words, one RANOPTIONS statement does not remember what
options were specified in previous RANOPTIONS statements. To have options persist over multiple
statements, use the NETOPTIONS statement.
connection-list
List of connections to randomize.
connection-list consists of 0 or more repetitions of:
wname -->
wname-2 where:
wname
is the layer ID, BIAS, or ALTITUDE
wname-2
is the layer ID
Options
RANDF=number
Specifies the degrees of freedom parameter for random numbers. See the
Randomization Options
and Default Parameters
table for values.
Default:
1
RANDIST=name
Specifies the type of distribution for random numbers. See the
Randomization Options and
Default Parameters
table for values.
Default:
NORMAL
RANDOM=integer
Specifies the random number seed.
Default:
0
RANLOC=number
Specifies the location parameter for random numbers. See the
Randomization Options and Default
Parameters
table for values.
RANSCALE=number
Specifies the scale parameter for random numbers. See the
Randomization Options and Default
Parameters
table for values.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The NEURAL Procedure
SAVE Statement
Writes weights to data set or a description of the network to a catalog entry.
Category Action Statement - affects the network or the data sets. Options
set in an action statement
affect only that statement.
NOTE: At least one option must be specified, but there is no single argument that is required.
SAVE OUTEST=
SAS-data-set
NETWORK=
screen-specification;
Options
Specify at least one:
NETWORK=screen-specification
Saves the definition of the entire network. screen-specification is the name of a catalog entry.
OUTEST= SAS-data-set
Saves the network weights in an output data set.
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The NEURAL Procedure
SCORE Statement
Creates an output data set containing predicted values and possibly other results such as residuals,
classifications, decisions, and assessment values.
Category Action Statement - affects the network or the data sets. Options set in an action statement
affect only that statement.
SCORE DATA=
SAS-data-set
OUT=
SAS-data-set
OUTFIT=
SAS-data-set
<
DUMMIES | NODUMMIES>
<
ROLE=
role-option>;
Required Arguments
OUT=SAS-data-set
Specifies the output data set that contains the outputs.
Options
DATA=SAS-data-set
Specifies the input data to be scored that contains inputs and, optionally, targets.
Default:
Defaults to the training data (DATA= in the PROC statement).
DUMMIES | NODUMMIES
Specifies whether to write dummy variables to the OUT= data set.
Default:
NODUMMIES
OUTFIT=SAS-data-set
Specifies the output data set that contains the fit statistics.
ROLE=role-option
Specifies the role of the DATA= data set. ROLE=role-option primarily affects which fit statistics
are computed and what their names and labels are. Role-option is one of the following:
TRAIN
Specifies that the DATA= data set is the training set. The data set must contain the target
variable.