Overview (OVERALS command)
OVERALS
performs
nonlinear canonical correlation analysis on two or more sets of variables.
Variables can have different optimal scaling levels, and no assumptions
are made about the distribution of the variables or the linearity
of the relationships.
Options
Optimal Scaling Levels. You can specify the level of optimal scaling at which you want to analyze each variable.
Number of Dimensions. You can specify how many dimensions OVERALS
should compute.
Iterations and Convergence. You can specify the maximum number of iterations and the value of a convergence criterion.
Display Output. The output can include all available statistics, only the default statistics, or only the specific statistics that you request. You can also control whether some of these statistics are plotted.
Saving Scores. You can save object scores in the active dataset.
Writing Matrices. You can write a matrix data file containing quantification scores, centroids, weights, and loadings for use in further analyses.
Basic Specification
- The basic specification
is command
OVERALS
, theVARIABLES
subcommand, theANALYSIS
subcommand, and theSETS
subcommand. By default,OVERALS
estimates a two-dimensional solution and displays a table listing optimal scaling levels of each variable by set, eigenvalues and loss values by set, marginal frequencies, centroids and weights for all variables, and plots of the object scores and component loadings.
Subcommand Order
- The
VARIABLES
subcommand,ANALYSIS
subcommand, andSETS
subcommand must appear in that order before all other subcommands. - Other subcommands can appear in any order.
Operations
- If the
ANALYSIS
subcommand is specified more than once,OVERALS
is not executed. For all other subcommands, if a subcommand is specified more than once, only the last occurrence is executed. -
OVERALS
treats every value in the range 1 to the maximum value that is specified onVARIABLES
as a valid category. To avoid unnecessary output, use theAUTORECODE
orRECODE
command to recode a categorical variable that has nonsequential values or that has a large number of categories. For variables that are treated as numeric, recoding is not recommended because the characteristic of equal intervals in the data will not be maintained (seeAUTORECODE
andRECODE
for more information).
Limitations
- String variables
are not allowed; use
AUTORECODE
to recode nominal string variables. - The data
must be positive integers. Zeros and negative values are treated as
system-missing, which means that they are excluded from the analysis.
Fractional values are truncated after the decimal and are included
in the analysis. If one of the levels of a categorical variable has
been coded 0 or some negative value, and you want to treat it as a
valid category, use the
AUTORECODE
orRECODE
command to recode the values of that variable. -
OVERALS
ignores user-missing value specifications. Positive user-missing values that are less than the maximum value that is specified on theVARIABLES
subcommand are treated as valid category values and are included in the analysis. If you do not want the category to be included, useCOMPUTE
orRECODE
to change the value to a value outside of the valid range. Values outside of the range (less than 1 or greater than the maximum value) are treated as system-missing and are excluded from the analysis. - If one variable in a set has missing data, all variables in that set are missing for that object (case).
- Each set must have at least three valid (non-missing, non-empty) cases.