R/criteria.cv.R

In amsantac/geospt: Geostatistical Analysis and Design of Optimal Spatial Sampling Networks

                                       ##########################
                                       ## Validacion Cruzada:  ##
                                       ##########################

assign("criteria.cv",
   function(m.cv){
      m.cv <- as.data.frame(m.cv)
      criterio.mx <- data.frame(matrix(0, ncol = 9))
      colnames(criterio.mx) <- c("MPE", "AKSE", "RMSPE", 
                                 "MSPE", "RMSSPE", "MAPPE", "CC", 
                                 "R2", "pseudoR2")
      resid.mean <- m.cv[, 3] - mean(m.cv[, 3])
      criterio.mx[, 1] <- mean(m.cv[, 4])
      criterio.mx[, 2] <- mean((m.cv[, 2])^0.5)
      criterio.mx[, 3] <- sqrt(sum((m.cv[, 4])^2)/nrow(m.cv))
      criterio.mx[, 4] <- mean(m.cv[, 5])
      criterio.mx[, 5] <- sqrt(sum((m.cv[, 5])^2)/nrow(m.cv))
      criterio.mx[, 6] <- mean(abs(m.cv[, 4]/m.cv[, 3]))
      criterio.mx[, 7] <- cor(m.cv[, 1], m.cv[, 3])
      criterio.mx[, 8] <- 1 - sum((m.cv[, 4])^2)/sum(resid.mean^2)
      criterio.mx[, 9] <- cor(m.cv[, 1], m.cv[, 3])^2
      criterio.mx
   })