Re: [R] kernel smoothing of disease rates at locations

Caveat: Both of the following claims are subject to verification by true experts, which I am not. But I believe: 1. If all values being smoothed are positive, then the smoother must be also. If there are negative values, this is no longer true, and your question needs much more detail to get an an

Re: [R] kernel smoothing of disease rates at locations

Is it possible to apply a kernel smoothing regression whose estimator or indeed the confidence intervals cannot take negative values or values greater than 1? Best regards, Ioanna -- View this message in context: http://r.789695.n4.nabble.com/kernel-smoothing-of-disease-rates-at-locations-tp79