The Average silhouette method is one of the most commonly used methods for determining the optimal number of clusters. It would useful to have the silhouette coefficient in the k-means and AHC outputs. An alternative method is the Gap statistic.

Make the whole, add in an ocx and or vcl component, with all api the hooks in it to set parameters. then we could plug this in a a3rd party application. even on mobile devices etc

The experiments conducted in agricultural research are generally being done based on pure statistical experimental designs like CRD, RBD, LSD, Factorial, ANOVA etc. for which a separate module is needed. All over the world, lot of agricultural exp...

There is a module on ecology; I don't see why not adding a new module on Evolution. This would not only complement with the Ecology module, but also broaden the scope to biology, medicine, paleontology students, researchers, teachers, etc.

When adding data from e.g. a 1H NMR spectrum (e.g. as bucket table) or similar spectra,being able to visualize the data as a spectrum and not only as numbers/peak integrals would be very nice. In this context, layering several spectra and statisti...

It would be very helpful if we could get 95% CI of the Sen's slope in Seasonal Mann-Kendall option. With my opinion, I think more option will be let the software much more flexible (such as calculate the seasonal Mann-Kendall test with more than 1...

In order to assess if a generated MaxDiff design has a good quality.

I suggest to add WoE method to XLSTAT and it would be a good solution for many area such as susceptibility map. The method has been widely used in GIS.

In GLMs add the binomial negative regression which fits sometimes better than Poisson with counting data