Proximity mapping: Data
With the Data tab, you can define one or more proximity sources based on variables in the dataset. Each source is created by selecting variables from the active dataset and specifying whether they represent proximity matrices or multivariate data from which proximities will be derived.
- Variables
- The Variables list displays all variables in the active dataset. The variables include numeric, ordinal, and, nominal variables eligible for use in the proximity computation. Select one or more variables and move them to Variables in Source.
- Variables in Source
- The Variables in Source lists the variables that are currently selected for the active proximity source. These variables define the structure of the source and are used to derive proximities (for multivariate data) or to specify a proximity matrix (if proximities are provided directly). Use the up or down arrow buttons to reorder the variables, or remove them as needed.
- Object Labels
- With Object Labels, you can specify a variable that provides labels for the objects (cases) to be displayed in the output plots and tables. Select a variable from the list and move it to Object Labels to assign it.
- Source definition
- To the right of the Variables in Source, the Source name and a Source
label fields define the internal identifier and descriptive label for the current proximity
source.The Source name must be unique identifier that is preceded by
$.
Use the following buttons to manage sources.
- Add Source creates a new proximity source from the variables in the Variables in Source pane.
- Change Source modifies the currently selected source using the updated variable selection.
- Remove Source deletes the the currently selected source from the list.
- Data Type
- Use the Data Type selector to specify how the selected variables are
interpreted. With the Data type group, you can indicate how proximities are
derived from the selected variables:
- Proximity data Use this option if your input is already a proximity matrix.
- Multivariate data (default) treats the selected variables as features of
the objects. PROXMAP derives proximities by computing distances between objects based on these variables.
- Euclidean: Proximities are based on unstandardized Euclidean distances.
- Standardized Euclidean: Variables are standardized before computing distances (default).
- Weights
- When Data type is set to Proximity data, an optional Proximity weights list becomes available. Here you can specify additional weight variables to be associated with the proximity values. These weights adjust the influence of each proximity in the optimization process.