replicate observation #40
Description
I am currently using the Bayesian Kernel Machine Regression (BKMR) method that you developed in my research. I have encountered an issue where the BKMR results appear to be anomalous when the dataset includes duplicate observations (i.e., instances where exposure values are identical across some observations).
Attached are two figures (Figure 1 and Figure 2) illustrating simple examples of this issue. I suspect that these duplicates might be influencing the model’s outputs in unexpected ways.
Could you please provide any insights or suggestions on how to address this issue? Any guidance would be greatly appreciated.
