Need Dev experienced in ALGLIB mql5 library to help debug my simple random forest test script

I am trying to use the ALGLIB library to work with decision trees / random forests

To learn how to use it, I created a small script to run a test on the iris flower data often used in some machine learning beginner example tutorials.

However, my script's precidtion output is always 0.

Documentation on the ALGLIB website is a little vague, so looking for help with debugging and education here.

Script attached.

Some notes:

• To my understanding, the training methods use the last column in the data set as the 'answer' or classification
• Because the library only accepts double values, I created an enum for the different iris flowers and changed the last col in the original iris data set, to the enum number.

Talk soon!