Operators, graphical objects that you can select from a palette and place in the canvas work area, form the nodes in a mining flow diagram. Each type of operator has a specific set of input and output data ports, one or more lists of data columns, and a number of properties that define exactly how each operator manipulates data. You can expand and collapse operators as you work with them to show more or less detail.
Input and output ports are depicted by blue dots on the operator. Ports on the left side of an operator represent input ports, those on the right side are output ports. Each type of operator has a different set of ports based on the function of the operator. Source operators, for example, only have output ports, transform operators generally have a number of input and output ports, and target operators only have input ports. Operator-level ports are visible in expanded and collapsed operators.
Most operator types have at least one list of columns. Many have input and output lists and some have multiple input and output lists. The column list in a source operator becomes populated with column names as the source table property is defined. Transform operator column lists are usually populated when you connect the operator to a defined source operator or to another transform operator with populated columns. Most operators have column lists with associated column-level connection ports. Column lists and column-level connections are only shown in an expanded operator.
Operator properties define how a specific operator behaves as part of a mining flow. In a source operator, the properties describe the location and details about the schema or format of the data source. Transform operator properties describe the specific details of each transformation step. The number and the complexity of properties depends on the type of transform operator.