plot_svm_jk: Plot the decision boundary of SVM model with 2 features
In ExabytE1337/Lori: Personal library of useful functions

Plot the decision boundary of SVM model with 2 features

plot_svm_jk(
  df,
  svm_model = NULL,
  grid = 50,
  tolerance = 0.005,
  color = c("#56b4e9", "#e69f00", "black"),
  plot_grid = T,
  plot_contour = T,
  plot_decision = T,
  fill_plot = F,
  fillmap = NULL,
  bins = 10,
  contour_color = "white",
  longer_legend = T,
  add_title = T,
  surface_plot = F,
  title = "SVM - Visualization",
  plot_data = T,
  ESL_theme = F,
  subtitle = T,
  ...
)

`df`	The data to be displayed in this layer.
`svm_model`	svm model from library e1071 or kernlab
`grid`	Size of the grid used for sampling the space. We have to evaluate the decision function f(x) in each point.
`tolerance`	when ploting the contours for the decision function f(x) = +-1, how close need the points to be to that line in order to be highlighted.
`color`	3 color vector. First two colors used for datapoints, third one used for the contours.
`plot_grid`	should the grid be plotted. Default value is TRUE.
`plot_contour`	should the contour for f(x) = +-1 be plotted. Default value is TRUE. This setting doesn't affect the contours for fill_plot.
`plot_decision`	should the decision boundary(the contour for f(x)=0) be plotted. Default value is TRUE.
`fill_plot`	should the plot be filled according to distance to the decision boundary. Classes have different shape. Filled shape means the point is a support vector. Default value is FALSE.
`fillmap`	colormap used for the fill_plot. Consider using diverging color pallete. Many available in the pals package.
`bins`	number of bins used for geom_contour in fill_plot.
`contour_color`	color used for contours in fill_plot.
`longer_legend`	should the legend be as long as the plot. Default value is TRUE.
`add_title`	should the title be added. Default value is TRUE.
`surface_plot`	should we plot the "decision" function f(x): R^2 -> R. Default value is FALSE.
`title`	string title for the plot.
`plot_data`	should the original data be plotted? Default value is TRUE.
`ESL_theme`	should x and y be without labels and ticks? Default value is FALSE.
`subtitle`	should default subtitle be added. Default value is TRUE.
`...`	additional paramaters when calling the persp_jk function.

data(df)
library(e1071)
model <- svm(label ~ x1 + x2, data = df, kernel = "radial", gamma = 1, scale = F, cost = 1.8)
plot_svm_jk(df,model, ESL_theme = T)

## filled plot
plot_svm_jk(df,model,fill_plot = T, plot_contour = F, plot_decision = F)

## surface plot
plot_svm_jk(df,model,surface_plot = T, theta = 300)