PLOT Subcommand (PROXSCAL command)

PLOT controls the display of plots. By default, PROXSCAL produces a scatterplot of object coordinates of the common space, the individual space weights, and the correlations between the independent variables (that is, equivalent to specifying COMMON, WEIGHTS, and CORRELATIONS).

  • Specifying a keyword overrides the default output and only output is generated for that keyword.
  • Duplicate keywords are silently ignored.
  • In case of contradicting keywords, only the last keyword is considered.
  • Inapplicable keywords (for example, stress with equal minimum and maximum number of dimensions on the CRITERIA subcommand) are silently ignored.
  • Multiple variable lists are allowed for TRANSFORMATIONS and RESIDUALS. For each variable list, a separate plot will be displayed.

NONE. No plots. PROXSCAL does not produce any plots.

STRESS. Stress plot. A plot is produced of stress versus dimensions. This plot is only produced if the maximum number of dimensions is larger than the minimum number of dimensions.

COMMON. Common space. A scatterplot matrix of coordinates of the common space is displayed.

WEIGHTS. Individual space weights. A scatterplot is produced of the individual space weights. This is only possible if one of the individual differences models is specified on the MODEL subcommand. For the weighted Euclidean model, the weights are printed in plots with one dimension on each axis. For the generalized Euclidean model, one plot is produced per dimension, indicating both rotation and weighting of that dimension. The reduced rank model produces the same plot as the generalized Euclidean model does but reduces the number of dimensions for the individual spaces.

INDIVIDUAL. Individual spaces. For each source specified on the variable list, the coordinates of the individual spaces are displayed in scatterplot matrices. This is only possible if one of the individual differences models is specified on the MODEL subcommand.

TRANSFORMATIONS. Transformation plots. Plots are produced of the original proximities versus the transformed proximities. On the variable list, the sources can be specified of which the plot is to be produced.

RESIDUALS. Residuals plots. The transformed proximities versus the distances are plotted. On the variable list, the sources can be specified of which the plot is to be produced.

VARIABLES. Independent variables. Transformation plots are produced for the independent variables specified on the variable list.

CORRELATIONS. Correlations. A plot of correlations between the independent variables and the dimensions of the common space is displayed.

Example

PROXSCAL VARIABLES=source01 TO source02 
  /TABLE=row_id BY col_id 
  /MODEL=WEIGHTED 
  /CRITERIA=DIMENSIONS(3) 
  /PLOT=COMMON INDIVIDUAL(source02).
  • Here, the syntax specifies a weighted Euclidean model with two sources in three dimensions.
  • COMMON produces a scatterplot matrix defined by dimensions 1, 2, and 3.
  • For the individual spaces, a scatterplot matrix with 3 dimensions is only produced for the individual space of source02.