JMP has several methods to explore outliers (Analyze>Screening>Explore outliers).
For a first data exploration, I am looking for a solution so the users can interactively filter all anomalous rows.
One simple approach is to conduct a KNN outlier analysis (unchecking the option of imputations and being able to change the default number)
![FN_0-1679518219072.png FN_0-1679518219072.png](https://community.jmp.com/t5/image/serverpage/image-id/51421i185C8CAD30B6B87B/image-size/medium?v=v2&px=400)
And save the distances to each NN, so a global score can be obtained (e.g., by adding all distances into one column).
![FN_1-1679518708197.png FN_1-1679518708197.png](https://community.jmp.com/t5/image/serverpage/image-id/51422i97839FFC4045877E/image-size/medium?v=v2&px=400)
How can automate these steps that have GUI actions in JMP 16?
Names Default To Here( 1 );
Clear Log();
dt = Open( "$SAMPLE_DATA/Bands Data.JMP" );
// Launch platform: Explore Outliers
Explore Outliers(
Y(
:proof cut, :viscosity, :caliper, :ink temperature, :humidity, :roughness,
:blade pressure, :varnish pct, :press speed, :ink pct, :solvent pct,
:ESA Voltage, :ESA Amperage, :wax, :hardener, :roller durometer,
:current density, :anode space ratio, :chrome content
)
);