Regression Line Plotter

This node takes a input data table plus a regression model of one target column against another value column and prepares a scatter plot joining the original data and a representation of the regression model output.

Since currently adding regression lines to scatter plots is not supported, the regression line is simulated with a sequence of dots. User can adjust the number of dots shown on the chart.

Updated Aug 2022, for multiple variables mode


NOTES for multiple variables:

Make sure to select the same target variable of the regression model and one of the predictor variables to plot.

The node will automatically average other predictors variables.

The first value will be used for categorical values (so better not using them in the predictor model).

This component only works with the output PMML models compatible with the Regression Predictor node. That is output ports from Linear or Polynomial Regression Learner nodes.

A big thanks to Paolo Tamagnini (https://www.knime.com/node/21589?) for great support in making suggestions/fixing this node!

Options

Select Input Features of the Model

Make sure all input predictors/features of the model are present in the list of columns on the left.

Input column

Please select column with regression input values ("X")

Target column

Please select column with target values(Y)

Maximum number of rows

Maximum number of rows from input table to be used. If the input is greater than this number we take the top 2.5k rows for the visualization.

Regression "line" number of dots

Enter the number of values predicted by the regression model to be plotte

Chart title

Enter chart title

Nodes

• GroupBy3 ×
• Math Formula (Variable)3 ×
• Color Manager2 ×
• Column Filter2 ×
• Column Rename (Regex)2 ×
• Show all 30 nodes

Extensions

• FeatureKNIME Base nodes
• FeatureKNIME Data Generation
• FeatureKNIME JavaScript Views
• FeatureKNIME Javasnippet
• FeatureKNIME Math Expression (JEP)
• Show all 6 modules

Links

• No links available