Union (RapidMiner Studio Core)

Synopsis

This operator builds the union of the input ExampleSets. The input ExampleSets are combined in such a way that attributes and examples of both input ExampleSets are part of the resultant union ExampleSet.

Description

The Union operator builds the superset of features of both input ExampleSets such that all regular attributes of both ExampleSets are part of the superset. The attributes that are common in both ExampleSets are not repeated in the superset twice, a single attribute is created that holds data of both ExampleSets. If the special attributes of both input ExampleSets are compatible with each other then only one special attribute is created in the superset which has examples of both the input ExampleSets. If special attributes of ExampleSets are not compatible, the special attributes of the first ExampleSet are kept. If both ExampleSets have any attributes with the same name, they should be compatible with each other; otherwise you will get an error message. This can be understood by studying the attached Example Process.

Input

  • example set 1 (IOObject)

    This input port expects an ExampleSet. It is the output of the Retrieve operator in the attached Example Process. The output of other operators can also be used as input. It is essential that meta data should be attached with the data for the input because attributes are specified in their meta data.

  • example set 2 (IOObject)

    This input port expects an ExampleSet. It is the output of the Retrieve operator in the attached Example Process. The output of other operators can also be used as input. It is essential that meta data should be attached with the data for the input because attributes are specified in their meta data.

Output

  • union (IOObject)

    The union of the input ExampleSets is delivered through this port.

Tutorial Processes

Union of the Golf and Golf-Testset data sets

In this process the 'Golf' data set and 'Golf-Testset' data set are loaded using the Retrieve operators. Breakpoints are inserted after the Retrieve operators so that you can have a look at the input ExampleSets. When you run the process, first you see the 'Golf' data set. As you can see, it has 14 examples. When you continue the process, you will see the 'Golf-Testset' data set. It also has 14 examples. Note that the meta data of both ExampleSets is almost the same. The Union operator is applied to combine these two ExampleSets into a single ExampleSet. The combined ExampleSet has all attributes and examples from the input ExampleSets, thus it has 28 examples. You can see that both input ExampleSets had the same number of attributes, same names and roles of attributes. This is why the Union ExampleSet also has the same number of attributes with the same names and roles. Here the Union operator behaves like the Append operator i.e. it simply combines examples of two ExampleSets with compatible meta data.

Union of the Golf and Iris data sets

In this process the 'Golf' data set and the 'Iris' data set are loaded using the Retrieve operators. Breakpoints are inserted after the Retrieve operators so that you can have a look at the input ExampleSets. When you run the process, first you see the 'Golf' data set. As you can see, it has 14 examples. When you continue the process, you will see the 'Iris' data set. It has 4 regular and 2 special attributes with 150 examples. Note that the meta data of both ExampleSets is very different. The Union operator is applied to combine these two ExampleSets into a single ExampleSet. The combined ExampleSet has all attributes and examples from the input ExampleSets, thus it has 164 (14+150) examples. Note that the 'Golf' data set has an attribute with label role: the 'Play' attribute. The 'Iris' data set also has an attribute with label role: the 'label' attribute. As these two label attributes are not compatible, only the label attribute of the first ExampleSet is kept. The examples of 'Iris' data set have null values in this attribute of the resultant Union ExampleSet.

Union of the Golf(with id attribute) and Iris data sets

In this process the 'Golf' data set and 'Iris' data set are loaded using the Retrieve operators. The Generate ID operator is applied on the Golf data set to generate nominal ids starting from id_1. Breakpoints are inserted before the Union operator so that you can have a look at the input ExampleSets. When you run the process, first you see the 'Golf' data set. As you can see, it has 14 examples. It has two special attributes. When you continue the process, you will see the 'Iris' data set. It has 4 regular and 2 special attributes with 150 examples. Note that the meta data of both ExampleSets is very different. The Union operator is applied to combine these two ExampleSets into a single ExampleSet. The combined ExampleSet has all attributes and examples from the input ExampleSets, thus it has 164 (14+150) examples. Note that the 'Golf' data set has an attribute with label role: the 'Play' attribute. The 'Iris' data set also has an attribute with label role: the 'label' attribute. As these two label attributes are not compatible, only the label attribute of the first ExampleSet is kept. The examples of the 'Iris' data set have null values in this attribute of the union ExampleSet. Also note that both input ExampleSets have id attributes. The names of these attributes are the same and they both have nominal values, thus these two attributes are compatible with each other. Thus a single id attribute is created in the resultant Union ExampleSet. Also note that the values of ids are not unique in the resultant ExampleSet.