The FACTEX Procedure


Summary of Functions

Table 7.1 to Table 7.4 classify the statements and options in PROC FACTEX by function.

Table 7.1: Summary of Options for Specifying the Design

Function

Statement

Option

Factor Specification

Factor names

FACTORS

$\mi{factor}_1 \ldots \mi{factor}_ f$

Number of levels

FACTORS

$\mi{factor}_1 \ldots \mi{factor}_ f$ / NLEV=q

Design Size Specification

(one of the following)

   

Number of runs

SIZE

DESIGN=n

Fraction of one full replicate

SIZE

FRACTION=h

Number of run indexing factors

SIZE

NRUNFACS=m

Minimum number of runs

SIZE

DESIGN=MINIMUM

   

or FRACTION=MAXIMUM

   

or NRUNFACS=MINIMUM

Block Specification

(one of the following)

   

Number of blocks

BLOCKS

NBLOCKS=b

Block size

BLOCKS

SIZE=k

Number of block pseudo-factors

BLOCKS

NBLKFACS=s

Minimum block size

BLOCKS

NBLOCKS=MAXIMUM

   

or SIZE=MINIMUM

   

or NBLKFACS=MAXIMUM

Model Specification

(one of the following)

   

Estimated effects

MODEL

ESTIMATE=(effects)

Estimated effects and

MODEL

ESTIMATE=(effects)

nonnegligible effects

 

NONNEG=(nonnegligible-effects)

Design resolution number

MODEL

RESOLUTION=r

Design with highest resolution

MODEL

RESOLUTION=MAXIMUM

Minimum aberration design

MODEL

EST=(…) <NONNEG=(…)> or RES=…

(up to dth-order interactions)

 

/ MINABS<(d)>


Table 7.2: Summary of Options for Searching the Design

Function

Statement

Option

Search for the Design

Allow maximum time of t seconds

PROC FACTEX

SECONDS=t or TIME=t

Limit the design searches

PROC FACTEX

NOCHECK


Table 7.3: Summary of Options for Replicating and Randomizing the Design

Function

Statement

Option

Replication

Replicate entire design c times

OUTPUT OUT=SAS-data-set

DESIGNREP=c

Replicate design for each point

OUTPUT OUT=SAS-data-set

DESIGNREP=SAS-data-set

in the data set

   

Replicate each point in design

OUTPUT OUT=SAS-data-set

POINTREP=p

p times

   

Replicate data set for each point

OUTPUT OUT=SAS-data-set

POINTREP=SAS-data-set

in the design

   

Randomization

Randomize the design

OUTPUT OUT=SAS-data-set

RANDOMIZE

Randomize the design but not the

OUTPUT OUT=SAS-data-set

RANDOMIZE NOVALRAN

assignment of factor levels

   

Specify seed number

OUTPUT OUT=SAS-data-set

RANDOMIZE $(u)$


Table 7.4: Summary of Options for Examining and Saving the Design

Function

Statement

Option

List the Design

Coded factor and block levels

EXAMINE

DESIGN

List the Design Characteristics

Alias structure

EXAMINE

ALIASING<(d)>

(up to dth-order interactions)

   

Confounding rules

EXAMINE

CONFOUNDING

Save the Design

Coded factor levels

OUTPUT OUT=SAS-data-set

 

Decoded factor levels

OUTPUT OUT=SAS-data-set

factor-name

(numeric type)

 

NVALS=(level1 $\dots $ levelq)

Decoded factor levels

OUTPUT OUT=SAS-data-set

factor-name

(character type)

 

CVALS=(’level1’ $\dots $ ’levelq)

Block variable name

OUTPUT OUT=SAS-data-set

BLOCKNAME=block-name

Decoded block levels

OUTPUT OUT=SAS-data-set

BLOCKNAME=block-name

(numeric type)

 

NVALS=(level1 $\dots $ levelb)

Decoded block levels

OUTPUT OUT=SAS-data-set

BLOCKNAME=block-name

(character type)

 

CVALS=(’level1’ $\dots $ ’levelb)