Categories

Versions

Transpose (AI Studio Core)

Synopsis

This operator transposes the input ExampleSet i.e. the current rows become columns of the output ExampleSet and current columns become rows of the output ExampleSet. This operator works very similar to the well known transpose operation for matrices.

Description

This operator transposes the input ExampleSet i.e. the current rows become the columns of the output ExampleSet and the current columns become the rows of the output ExampleSet. In other words every example or row becomes a column with attribute values and each attribute column becomes an example row. This operator works very similar to the well known transpose operation for matrices. The transpose of a transpose of a matrix is same as the original matrix, but the same rule cannot be applied here because the types of the original ExampleSet and the transpose of the transpose of an ExampleSet may be different.

If an id attribute is part of the input ExampleSet, the ids will become the names of the new attributes. The names of the old attributes will be transformed into the id values of a new id attribute. All other new attributes will have regular role after the transformation. You can use the Set Role operator after the transpose operator to assign roles to new attributes.

If all old attributes have the same value type, all new attributes will have the same value type. If at least one nominal attribute is part of the input ExampleSet, the type of all new attributes will be nominal. If the old attribute values were all mixed numbers, the type of all new attributes will be real. This operator produces a copy of the data in the main memory. Therefore, it should not be used on very large data sets.

Input

  • example set input (Data table)

    This input port expects an ExampleSet. It is output of the Retrieve operator in the attached Example Process.

Output

  • example set output (Data table)

    The transpose of the input ExampleSet is output of this port.

  • original (Data table)

    The ExampleSet that was given as input is passed without changing to the output through this port. This is usually used to reuse the same ExampleSet in further operators or to view the ExampleSet in the Results Workspace.

Tutorial Processes

Different scenarios of Transpose

There are four different cases in this Example Process:

Case 1: The 'Golf' data set is loaded using the Retrieve operator. A breakpoint is inserted here so that you can have a look at the ExampleSet before application of the Transpose operator. You can see that the 'Golf' data set has no id attribute. The types of attributes are different including attributes of nominal type. Press the Run button to continue. Now the Transpose operator is applied on the 'Golf' data set. A breakpoint is inserted here so that you can see the ExampleSet after the application of the Transpose operator. Here you can see that a new attribute with id role has been created. The values of the new id attribute are the names of the old attributes. New attributes are named in a general format like 'att_1', 'att_2' etc because the input ExampleSet had no id attribute. The type of all new attributes is nominal because there were attributes with different types including at least one nominal attribute in the input ExampleSet.

Case 2: The 'Iris' data set is loaded using the Retrieve operator. A breakpoint is inserted here so that you can have a look at the ExampleSet before application of the Transpose operator. You can see that the 'Iris' data set has an id attribute. The types of attributes are different including attributes of nominal type. Press the Run button to continue. Now the Transpose operator is applied on the 'Iris' data set. A breakpoint is inserted here so that you can see the ExampleSet after the application of the Transpose operator. Here you can see that a new attribute with id role has been created. The values of the new id attribute are the names of the old attributes. The ids of the old ExampleSet become names of the new attributes. The type of all new attributes is nominal because there were attributes with different types including at least one nominal attribute in the input ExampleSet.

Case 3:The 'Market-Data' data set is loaded using the Retrieve operator. A breakpoint is inserted here so that you can have a look at the ExampleSet before application of the Transpose operator. You can see that the 'Market-Data' data set has no special attributes. The type of all attributes is integer. Press the Run button to continue. Now the Transpose operator is applied on the 'Market-Data' data set. A breakpoint is inserted here so that you can see the ExampleSet after the application of the Transpose operator. Here you can see that a new attribute with id role has been created. Values of the new id attribute are the names of the old attributes. New attributes are named in a general format like 'att_1', 'att_2' etc because the input ExampleSet had no id attribute. The Type of all new attributes is real because there were attributes with mixed numbers type in the input ExampleSet.

Case 4:The 'Golf-Testset' data set is loaded using the Retrieve operator. A breakpoint is inserted here so that you can have a look at the ExampleSet before application of the Transpose operator. The Transpose operator is applied on the 'Golf-Testset' data set. Then the Transpose operator is applied on the output of the first Transpose operator. Note that the types of the attributes of the original ExampleSet and the Transpose of the Transpose of the original data set are different.