Linear correlation (matrices and scatter plots)

The component calculates the correlation between two or multiple variables (continuous or categorical features) and displays the correlation measures and matrices accordingly.
For the continuous variables a scatter plot can be displayed for selected variables.
A correlation filter with threshold 0.7 removes correlated columns, joins the remaining columns with the nominal columns of the data set and provides the result at the output. The threshold can be changed in Correlation Filter node.

Input ports

1. Type: Flow Variable
   Port 1
   Flow variables named Default_Var_nominal and Default_Var_cont. They can be created by the component: Split features & create default flow variables.
2. Type: Table
   Port 2
   Data table with categorical nominal value columns.
3. Type: Table
   Port 3
   Data table with continuous value columns.

Output ports

1. Type: Table
   Port 1
   Data table without the correlated columns.

External resources

• Scatter-plot icons created by Freepik - Flaticon