R/plot_linreg.R

In JS2387/Group8LinearRegressionFJ: Runs regression and fits linear models

#' @title Generic Method for plot()
#' @param x object containing model data to be used for plotting
#' @param y the y coordinates for plotting
#' @param ... other parameters for plotting
#' @import  ggplot2
#' @export
plot <- function (x, y, ...) UseMethod("plot")
#' @title linreg Method for plot()
#' @param x object containing model data to be used for plotting
#' @param y the y coordinates for plotting
#' @param ... other parameters for plotting
#' @import  ggplot2 stats
#' @export
plot.linreg <- function (x, y, ...) {
                          # Create the DF used for plotting
                          #library(latex2exp)
                          #library(ggplot2)
                          error_hat <- x$Residuals
                          y_hat <- x$Fittedvalues
                          df_1 <- data.frame(error_hat , y_hat)

                          # Smooth variable to create the desired plot
                          smoothed <- stats::lowess(x = y_hat, y = error_hat)

                          # Create the first plot
                          plot_res_vs_fitted <- ggplot2::ggplot(df_1 , aes(x=y_hat , y=error_hat) )+
                            geom_point(shape = 1 , size=4) +
                            geom_line(aes(x = smoothed$x, y = smoothed$y), color = "red") +
                            ggtitle("Residuals vs Fitted") +
                            theme(plot.title = element_text(hjust = 0.5))+
                            ylab("Residuals")+
                            xlab("Fitted values")+
                            theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),
                                  panel.background = element_blank(), axis.line = element_line(colour = "black")) +
                            theme(plot.margin=unit(c(2,2,2.5,1.2),"cm"))

                          #return(plot_res_vs_fitted)

                          # SECOND PLOT THING SCALE LOCATION

                          # Calculate standardised residuals

                          standard_res <- error_hat / sd(error_hat)
                          standard_res_sqrt <- sqrt(abs(standard_res))
                          smoothed_2 <- stats::lowess(x = y_hat, y = standard_res_sqrt)


                          #Second plotting DF

                          df_2 <- data.frame(y_hat , standard_res_sqrt )


                          plot_scale_location  <- ggplot2::ggplot(df_2 , aes(x=y_hat , y= standard_res_sqrt ))+
                            geom_point(shape = 1 , size=4) +
                            geom_line(aes(x = smoothed_2$x, y = smoothed_2$y), color = "red") +
                            ggtitle("Scale-Location") +
                            theme(plot.title = element_text(hjust = 0.5))+
                            ylab("Sqrt(Standardised Residuals)")+
                            xlab("Fitted values")+
                            theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(),
                                  panel.background = element_blank(), axis.line = element_line(colour = "black")) +
                            theme(plot.margin=unit(c(2,2,2.5,1.2),"cm"))

                          list(plot_res_vs_fitted, plot_scale_location)

                          #return(plot_scale_location)

}