R package providing basic functionalities to use projection pursuit regression (PPR) in context of regional frequency analysis.
Regression method can be used to predict at ungauged location a hydrological variable Y based on the avalaible site characteristics
A Ridge function has the form g(a'x) where x is vector of explanatory variable,a is a direction, or a unitary vector of coefficients and g is a nonparametric function. A PPR model is regression model defines by several Ridge terms
E(Y|x) = u + g_1(a_1'x) + g_2(a_2'x) + ... + error
and can be seen as a generalized additive model with estimated linear predictor a'x.
This packages provide wrapping functions and graphical tools to create a gam object of the well known mgcv R packages, with additional direction.
For further information please consult the R documentation.
The package is not in the CRAN repository. It can be download from the present github page and built locally.
version v0.2: correct issues of v0.1
version v0.1: first complete version
This package still needs further testing and an improved documentation. There will be frequent commits. At this moments the program should not be considered as finished and must be used carefully.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.