Singular Value Decomposition

This transform reduces the dimensionality by Singular Value Decomposition method. It is used in feature engineering and is applied only to the numerical data.

tags: [“Data Preparation”]

Parameters

The table gives a brief description about each parameter in Shuffle Dataset transform.

Name:

By default, the transform name is populated. You can also add a custom name for the transform.

Input Dataset:

The file name of the input dataset. You can select this from the drop-down list to reduce by singular value decomposition. (Required: True, Multiple: False)

Output Dataset:

The file name of the SVD reduced Dataset. (Required: True, Multiple: False)

Number of components:

The total number of columns to which SVD must be applied. The value should be less than the number of columns in the dataset. (Required: True, Multiple: False, Datatypes: [“LONG”], Options: [‘CONSTANT’])

Target:

The column that is the target variable in the dataset. This is the variable on which predictions are made. (Required: True, Multiple: False, Datatypes: [“STRING”], Options: [‘FIELDS’], Datasets: [‘df’])

The sample input for this transform looks as shown in the screenshot:

The output after running the Singular Value Decomposition transform on the dataset appears as below:

How to use it in Notebook

template=TemplateV2.get_template_by('Singular Value Decomposition')

recipe_Singular_Value_Decomposition= project.addRecipe([car_data, employee_data, temperature_data, only_numeric], name='Singular Value Decomposition')

transform=Transform()
transform.templateId = template.id
transform.name='Singular Value Decomposition'
transform.variables = {
'input_dataset':'only_numeric',
'output_dataset':'svd',
'n':1,
'target':"Age"}
recipe_Singular_Value_Decomposition.add_transform(transform)
recipe_Singular_Value_Decomposition.run()

Requirements

scikit-learn pandas