Assigning Data to Roles
To run the Correspondence Analysis task, you must
select an input data source. To filter the input data source, click 
.
In the Correspondence Analysis task, you can choose the data layout:
raw data or table data. The Roles options
differ depending on the data layout.
.
In the Correspondence Analysis task, you can choose the data layout:
raw data or table data. The Roles options
differ depending on the data layout.
Simple Correspondence Analysis of Raw Data
These options are available
if you select a data layout of raw data and the simple correspondence
analysis.
To run a simple correspondence
analysis of raw data, you must assign variables to the Row
variables and Column variables roles.
|
Option Name
|
Description
|
|---|---|
|
Roles
|
|
|
Row variables
|
specifies the values
to use to construct the rows of the contingency table.
|
|
Column variables
|
specifies the values
to use to construct the columns of the contingency table.
|
|
Treat missing
values as a distinct level
|
specifies whether to
include missing values in the analysis.
|
|
Additional Roles
|
|
|
Supplementary
row variables
|
specifies variables
to represent as points in the joint row and column space. These variables
are not used in determining the locations of the other active row
and column points of the contingency table. Supplementary observations
on supplementary variables are ignored in simple correspondence analysis
but are needed to compute the squared cosines for multiple correspondence
analysis.
Note: This role is not available
if you assign row variables and select The factorial combinations
of levels of the row variables appear as table row from
the Create the contingency table rows drop-down
list.
|
|
Supplementary
column variables
|
specifies variables
to represent as points in the joint row and column space. These variables
are not used in determining the locations of the other active row
and column points of the contingency table. Supplementary observations
on supplementary variables are ignored in simple correspondence analysis
but are needed to compute the squared cosines for multiple correspondence
analysis.
Note: This role is not available
if you assign column variables and select The factorial
combinations of levels of the column variables appear as table columns option
from the Create the contingency table columns drop-down
list.
|
|
Weight variable
|
specifies weights for
each observation and indicates supplementary observations for simple
correspondence analyses.
|
|
Group analysis
by
|
creates separate analyses
of observations in groups that are defined by the BY variables.
|
Multiple Correspondence Analysis of Raw Data
These options are available
if you select a data layout of raw data and the multiple correspondence
analysis.
To run a multiple correspondence
analysis of raw data, you must assign a variable to the Column
variables role.
|
Option Name
|
Description
|
|---|---|
|
Roles
|
|
|
Column variables
|
specifies the values
to use to construct the columns of the contingency table. You must
assign at least two variables to this role.
|
|
Treat missing
values as a distinct level
|
specifies whether to
include missing values in the analysis.
|
|
Additional Roles
|
|
|
Supplementary
column variables
|
specifies variables
to represent as points in the joint row and column space. These variables
are not used in determining the locations of the other active row
and column points of the contingency table. Supplementary observations
on supplementary variables are ignored in simple correspondence analysis
but are needed to compute the squared cosines for multiple correspondence
analysis.
|
|
Weight variable
|
specifies weights for
each observation and indicates supplementary observations for simple
correspondence analyses.
|
|
Group analysis
by
|
creates separate analyses
of observations in groups that are defined by the BY variables.
|
Simple Correspondence Analysis of Table Data
These options are available
if you select a data layout of table data and the simple correspondence
analysis.
To run a simple correspondence
analysis of table data, you must assign at least two variables to
the Columns of the contingency table role.
|
Option Name
|
Description
|
|---|---|
|
Roles
|
|
|
Columns
of the contingency table
|
reads an existing contingency
table, binary indicator matrix, fuzzy-coded indicator matrix, or Burt
table, rather than raw data.
|
|
Additional Roles
|
|
|
Supplementary
columns of the contingency table
|
specifies variables
to represent as points in the joint row and column space. These variables
are not used in determining the locations of the other active row
and column points of the contingency table. Supplementary observations
on supplementary variables are ignored in simple correspondence analysis
but are needed to compute the squared cosines for multiple correspondence
analysis.
|
|
Label contingency
table rows
|
labels the rows of the
tables with the values of this variable and includes this variable
in the output data set.
|
|
Weight variable
|
specifies weights for
each observation and indicates supplementary observations for simple
correspondence analyses for table data.
|
|
Group analysis
by
|
creates separate analyses
of observations in groups that are defined by the BY variables.
|
Multiple Correspondence Analysis of Table Data
These options are available
if you select a data layout of table data and the multiple correspondence
analysis.
To run the Correspondence
Analysis task, you must assign variables to the Columns
of the Burt table role. A Burt table is a symmetric matrix
of crosstabulations among several categorical variables.
|
Option Name
|
Description
|
|---|---|
|
Roles
|
|
|
Columns
of the Burt table
|
specifies the variables
to use in the analysis.
|
|
Number of
variables used to create the Burt table
|
specifies the number
of classification variables to use to create the table.
|
|
Additional Roles
|
|
|
Supplementary
columns of the Burt table
|
specifies variables
to represent as points in the joint row and column space. These variables
are not used in determining the locations of the other active row
and column points of the contingency table. Supplementary observations
on supplementary variables are ignored in simple correspondence analysis
but are needed to compute the squared cosines for multiple correspondence
analysis.
|
|
Group analysis
by
|
creates separate analyses
of observations in groups that are defined by the BY variables.
|
Copyright © SAS Institute Inc. All rights reserved.