Test the accuracy of the GMM in adaptive threshold calculation. E.g., try different kernel representations to see if the Gaussian mixture always fits well or not... for example, without removing the zeros, the fit seems much worse, but I don't understand if it's a bad rescaling or just a bad fit. Try different kernels and, importantly, show the true histogram of the data (no smoothing kernel).
A different model (e.g., exponential+normal) would probably be better suited to the distribution of the data.
In addition, varargs (...) do not work in count_density r4tcpl function. See the related r4tcpl issue
Test the accuracy of the GMM in adaptive threshold calculation. E.g., try different kernel representations to see if the Gaussian mixture always fits well or not... for example, without removing the zeros, the fit seems much worse, but I don't understand if it's a bad rescaling or just a bad fit. Try different kernels and, importantly, show the true histogram of the data (no smoothing kernel).
A different model (e.g., exponential+normal) would probably be better suited to the distribution of the data.
In addition, varargs (
...) do not work incount_densityr4tcpl function. See the related r4tcpl issue