- ML Inspector version: 0.0.17
- Python version: 3.6.1
- Operating System: Ubuntu 17.04
Description
Feature dependence plots do not show which indices of the input test data they are mapped to. The indices must be matched manually which gets increasingly challenging with the number of plotted lines. Ideally, the indices and their corresponding colors would be shown in a legend. For purely aesthetic reasons, the dots representing the original values of the test data should share the same color as the lines they sit on.
What I Did
mlitest= X_testE.sample(500)
mxr=mli.ModelXRay(gbmodelE,mlitest)
plt.show()
Description
Feature dependence plots do not show which indices of the input test data they are mapped to. The indices must be matched manually which gets increasingly challenging with the number of plotted lines. Ideally, the indices and their corresponding colors would be shown in a legend. For purely aesthetic reasons, the dots representing the original values of the test data should share the same color as the lines they sit on.
What I Did