GWR collinearity diagnostics, prediction, F tests and GW summary statistics

GwrCollinearity: local weighted correlations, VIFs, condition numbers and variance-decomposition proportions, as GWmodel::gwr.collin.diagno. GwrPredict: GWR prediction with prediction variance, as GWmodel::gwr.predict. GwrFTests: Leung, Mei and Zhang's F1 to F4 tests (method = "gwmodel" reproduces GWmodel 2.4's simplifications). GwSummary: geographically weighted means, variances, skewness, CVs, covariances and Pearson and Spearman correlations, as GWmodel::gwss. Identical to the Python arm morie.fn.gwrext.

Usage

GwrCollinearity(X, coords, bw, kernel = "bisquare", adaptive = FALSE)

GwrPredict(
  y,
  X,
  coords,
  X0,
  coords0,
  bw,
  kernel = "bisquare",
  adaptive = FALSE
)

GwrFTests(
  y,
  X,
  coords,
  bw,
  kernel = "bisquare",
  adaptive = FALSE,
  method = "leung"
)

GwSummary(X, coords, bw, kernel = "bisquare", adaptive = FALSE)

Arguments

X

Design (intercept first) or data columns (GwSummary).

coords

Two-column coordinates.

bw

Bandwidth.

kernel

Kernel (as GWRKernelWeights).

adaptive

Adaptive bandwidth.

y

Response.

X0, coords0

Prediction design and coordinates.

method

"leung" or "gwmodel".

Value

List.

References

Wheeler, D. C. (2007). Diagnostic tools and a remedial method for collinearity in geographically weighted regression. Environment and Planning A 39, 2464-2481.

Leung, Y., Mei, C.-L. and Zhang, W.-X. (2000). Statistical tests for spatial nonstationarity based on the geographically weighted regression model. Environment and Planning A 32, 9-32.

Brunsdon, C., Fotheringham, A. S. and Charlton, M. (2002). Geographically weighted summary statistics. Computers, Environment and Urban Systems 26, 501-524.

Examples

X <- cbind(1, 0:4)
GwrPredict(c(1, 3, 5, 7, 9), X, cbind(0:4, 0), rbind(c(1, 2.5)), rbind(c(2.5, 0)), 100)$prediction
#> [1] 6
GwSummary(cbind(1:3, c(2, 4, 6)), cbind(0:2, 0), 100)$corr
#> [[1]]
#> [1] 1
#> 
#> [[2]]
#> [1] 1
#> 
#> [[3]]
#> [1] 1
#>