R/ctree_LUR.R

In mengluchu/APMtools: statistical air pollution mapping tools

#' Regression tree for LUR (using ctree), mainly for visualisation purpose

#' @param  variabledf the dataframe containing predictors and dependent variable
#' @param y_varname  name of the dependent variable.
#' @param training the index for the rows used for training.
#' @param grepstring the variable/column names of predictors in Lasso, grepl stlye, e.g. 'ROAD|pop|temp|wind|Rsp|OMI|eleva|coast'
#' @return plot the tree and return the ctree object
#' @export
ctree_LUR = function(variabledf, vis1 = T, y_varname = c("day_value", "night_value", "value_mean"), training, test, grepstring, ...) {

    prenres = paste(y_varname, "|", grepstring, sep = "")
    pre_mat = subset_grep(variabledf[training, ], prenres)


    y_test = variabledf[test, y_varname]
    x_test = variabledf[test, ]

    formu = as.formula(paste(y_varname, "~.", sep = ""))

    cf = ctree(formu, data = pre_mat)


    pre_rf <- predict(cf, newdata = x_test)
    # rf_residual <- pre_rf - rdf_test$NO2


    if (vis1) {
        print(plot(cf, fitmean = T))  #ctree party
    }
    # return(cf) cf2=rpart(formu, data=pre_mat)
    return(error_matrix(y_test, pre_rf))
}