Type Node
Field properties can be specified in a source node or in a separate Type node. The functionality is similar in both nodes. The following properties are available:
- Field Double-click any field name to specify value and field labels for data in IBM® SPSS® Modeler. For example, field metadata imported from IBM SPSS Statistics can be viewed or modified here. Similarly, you can create new labels for fields and their values. The labels that you specify here are displayed throughout IBM SPSS Modeler depending on the selections you make in the stream properties dialog box.
- Measurement This is the measurement level, used to
describe characteristics of the data in a given field. If all of the details of a field are known,
it is called fully instantiated. For more information, see Measurement levels. Note: The measurement level of a field is different from its storage type, which indicates whether the data are stored as strings, integers, real numbers, dates, times, timestamps, or lists.
- Values This column enables you to specify options for
reading data values from the data set, or use the Specify option to specify
measurement levels and values in a separate dialog box. You can also choose to pass fields without
reading their values. For more information, see Data Values. Note: You cannot amend the cell in this column if the corresponding Field entry contains a list.
- Missing Used to specify how missing values for the
field will be handled. For more information, see Defining Missing Values. Note: You cannot amend the cell in this column if the corresponding Field entry contains a list.
- Check In this column, you can set options to ensure
that field values conform to the specified values or ranges. For more information, see Checking Type Values. Note: You cannot amend the cell in this column if the corresponding Field entry contains a list.
- Role Used to tell modeling nodes whether fields will be Input (predictor fields) or Target (predicted fields) for a machine-learning process. Both and None are also available roles, along with Partition, which indicates a field used to partition records into separate samples for training, testing, and validation. The value Split specifies that separate models will be built for each possible value of the field. For more information, see Setting the field role.
Several other options can be specified using the Type node window:
- Using the tools menu button, you can choose to Ignore Unique Fields once a Type node has been instantiated (either through your specifications, reading values, or running the stream). Ignoring unique fields will automatically ignore fields with only one value.
- Using the tools menu button, you can choose to Ignore Large Sets once a Type node has been instantiated. Ignoring large sets will automatically ignore sets with a large number of members.
- Using the tools menu button, you can choose to Convert Continuous Integers To Ordinal once a Type node has been instantiated. See the topic Converting Continuous Data for more information.
- Using the tools menu button, you can generate a Filter node to discard selected fields.
- Using the sunglasses toggle buttons, you can set the default for all fields to Read or Pass. The Types tab in the source node passes fields by default, while the Type node itself reads values by default.
- Using the Clear Values button, you can clear changes to field values made in this node (non-inherited values) and reread values from upstream operations. This option is useful for resetting changes that you may have made for specific fields upstream.
- Using the Clear All Values button, you can reset values for all fields read into the node. This option effectively sets the Values column to Read for all fields. This option is useful to reset values for all fields and reread values and types from upstream operations.
- Using the context menu, you can choose to Copy attributes from one field to another. See the topic Copying Type Attributes for more information.
- Using the View unused field settings option, you can view type settings for fields that are no longer present in the data or were once connected to this Type node. This is useful when reusing a Type node for datasets that have changed.