The function returns a vector of weights using the tricube scheme:

$$w_{ij}(g) = (1 - (d_{ij}/d)^3)^3 $$ if \(d_{ij} <= d\) else \(w_{ij}(g) = 0\), where \(d_{ij}\) are the distances between the observations and \(d\) is the distance at which weights are set to zero.

gwr.tricube(dist2, d)

Arguments

dist2

vector of squared distances between observations

d

distance at which weights are set to zero

Value

matrix of weights.

References

Fotheringham, A.S., Brunsdon, C., and Charlton, M.E., 2000, Quantitative Geography, London: Sage; C. Brunsdon, A.Stewart Fotheringham and M.E. Charlton, 1996, "Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity", Geographical Analysis, 28(4), 281-298; http://gwr.nuim.ie/

Author

Roger Bivand Roger.Bivand@nhh.no

See also

Examples

plot(seq(-10,10,0.1), gwr.tricube(seq(-10,10,0.1)^2, 6.0), type="l")