PLOT Subcommand (PRINCALS command)
PLOT
can
be used to produce plots of category quantifications, object scores,
and component loadings.
- If
PLOT
is not specified, plots of the object scores and the quantifications are produced. - No plots are produced for a one-dimensional solution.
PLOT
is
followed by one or more of the following keywords:
LOADINGS. Plots of the component loadings of SNOM, ORDI, and NUME variables.
OBJECT. Plots of the object scores.
QUANT. Plots of the category quantifications for MNOM variables and plots of the single-category coordinates for SNOM, ORDI, and NUME variables.
DEFAULT. QUANT and OBJECT.
ALL. All available plots.
NONE. No plots.
- The keywords
OBJECT
andQUANT
can each be followed by a variable list in parentheses to indicate that plots should be labeled with these variables. ForQUANT
, the variables must be specified on both theVARIABLES
andANALYSIS
subcommands. ForOBJECT
, the variables must be specified onVARIABLES
but need not appear on theANALYSIS
subcommand. This means that variables not included in the computations can still be used to labelOBJECT
plots. If the variable list is omitted, only the default plots are produced. - Object scores plots labeled with variables that appear
on the
ANALYSIS
subcommand use category labels corresponding to all categories within the defined range. Objects in a category that is outside the defined range are labeled with the label corresponding to the next category greater than the defined maximum category. - Object scores plots labeled with variables not included
on the
ANALYSIS
subcommand use all category labels, regardless of whether or not the category value is inside the defined range. - All of the keywords except
NONE
can be followed by an integer in parentheses to indicate how many characters of the variable or value label are to be used on the plot. (If you specify a variable list afterOBJECT
orQUANT
, you can specify the value in parentheses after the list.) The value can range from 1 to 20. If the value is omitted, 12 characters are used. Spaces between words count as characters. - The
LOADINGS
plots and one of theQUANT
plots use variable labels; all other plots that use labels use value labels. - If a variable label is missing, the variable name is used for that variable. If a value label is missing, the actual value is used.
- You should make sure that your variable and value labels are unique by at least one letter in order to distinguish them on the plots.
- When points overlap, the points involved are described in a summary following the plot.
Example
PRINCALS VARIABLES COLA1 (4) COLA2 (4) COLA3 (4) COLA4 (2)
/ANALYSIS COLA1 COLA2 (SNOM) COLA3 (ORDI) COLA4 (ORDI)
/PLOT OBJECT(COLA4).
- Four variables are included in the analysis.
-
OBJECT
requests a plot of the object scores labeled with the values of COLA4. Any object whose COLA4 value is not 1 or 2 is labeled 3 (or the value label for category 3, if defined).
Example
PRINCALS VARIABLES COLA1 (4) COLA2 (4) COLA3 (4) COLA4 (2)
/ANALYSIS COLA1 COLA2 (SNOM) COLA3 (ORDI)
/PLOT OBJECT(COLA4).
- Three variables are included in the analysis.
-
OBJECT
requests a plot of the object scores labeled with the values of COLA4, a variable not included in the analysis. Objects are labeled using all values of COLA4.
In addition to the plot keywords, the following can be specified:
NDIM. Dimension
pairs to be plotted. NDIM
is followed by a pair of values in parentheses. If NDIM
is not specified, plots are produced
for dimension 1 versus dimension 2.
- The first value indicates the dimension that is plotted against all higher dimensions. This value can be any integer from 1 to the number of dimensions minus 1.
- The second value indicates the highest dimension to be used in plotting the dimension pairs. This value can be any integer from 2 to the number of dimensions.
- The keyword
ALL
can be used instead of the first value to indicate that all dimensions are paired with higher dimensions. - The keyword
MAX
can be used instead of the second value to indicate that plots should be produced up to, and including, the highest dimension fit by the procedure.
Example
PRINCALS COLA1 COLA2 COLA3 COLA4 (4)
/PLOT NDIM(1,3) QUANT(5).
- The
NDIM(1,3)
specification indicates that plots should be produced for two dimension pairs—dimension 1 versus dimension 2 and dimension 1 versus dimension 3. -
QUANT
requests plots of the category quantifications. The (5) specification indicates that the first five characters of the value labels are to be used on the plots.
Example
PRINCALS COLA1 COLA2 COLA3 COLA4 (4)
/PLOT NDIM(ALL,3) QUANT(5).
- This plot is the same as above except for the
ALL
specification followingNDIM
. This indicates that all possible pairs up to the second value should be plotted, soQUANT
plots will be produced for dimension 1 versus dimension 2, dimension 2 versus dimension 3, and dimension 1 versus dimension 3.