Skip to contents

Return and plot accumulated kernel

get_kernel.Rd

Return and plot the accumulated kernel, consisting of one or multiple windows

Usage

get_kernel(param, mix = NULL, kernel_type = "single", weighted = TRUE)

# S3 method for SWR
coef(object, ...)

plot_kernel(
  list = NULL,
  param = NULL,
  mix = NULL,
  kernel_type = "single",
  weighted = TRUE,
  xlim = NULL,
  include_text = TRUE,
  scales = "fixed"
)

# S3 method for SWR
plot(x, type = "kernel", ...)

Arguments

param

a matrix with 2 columns, see details

mix

a vector of mixing parameters (beta), see details

kernel_type

a String indicating whether (a) all single windows (option: "single"), or (b) the combined kernel (option "combined")

weighted

if TRUE, windows are weighted with mix parameters; default: TRUE

object

an SWR model object created using trainSWR

...

further plotting parameters, see plot_kernel or plot_prediction, respectively

list

a list containing multiple trained SlidingWindowReg models

xlim

vector with lower / upper bound of the x axis to print the kernel

include_text

if TRUE, plots will be annotated

scales

whether scales should be shared across multiple sub-plots ("fixed"), or free ("free")

x

an SWR model object

type

either "kernel" to produce a plot of the window kernels, or "prediction" to plot the model predictions, see plot_prediction

Functions

  • get_kernel(): return the kernel of an SWR model

  • coef(SWR): return the kernel of an SWR model

  • plot_kernel(): plot the kernel of an SWR model

  • plot(SWR): plot the kernel of an SWR model

Examples

param <- cbind(
  delta = c(0, 10),
  sigma = c(2, 3))
mix <- rep(1, ncol(param))
mod <- createSWR(param = param, mix = mix)

get_kernel(param = param, mix = mix)
#>                                                                             
#> kernel1 0.000000000 0.000000000 0.000000000 0.00000000 0.00000000 0.00000000
#> kernel2 0.001534648 0.003912432 0.008934251 0.01827455 0.03348232 0.05495003
#>                                                                                
#> kernel1 0.00000000 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000 0.00000000
#> kernel2 0.08078044 0.1063732 0.1254721 0.1325721 0.1254721 0.1063732 0.08078044
#>                  -6         -5         -4          -3          -2          -1
#> kernel1 0.004017088 0.01545604 0.04653623 0.109659595 0.202259910 0.292021009
#> kernel2 0.054950035 0.03348232 0.01827455 0.008934251 0.003912432 0.001534648
#>                 0
#> kernel1 0.3300501
#> kernel2 0.0000000
coef(mod)
#>                                                                             
#> kernel1 0.000000000 0.000000000 0.000000000 0.00000000 0.00000000 0.00000000
#> kernel2 0.001534648 0.003912432 0.008934251 0.01827455 0.03348232 0.05495003
#>                                                                                
#> kernel1 0.00000000 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000 0.00000000
#> kernel2 0.08078044 0.1063732 0.1254721 0.1325721 0.1254721 0.1063732 0.08078044
#>                  -6         -5         -4          -3          -2          -1
#> kernel1 0.004017088 0.01545604 0.04653623 0.109659595 0.202259910 0.292021009
#> kernel2 0.054950035 0.03348232 0.01827455 0.008934251 0.003912432 0.001534648
#>                 0
#> kernel1 0.3300501
#> kernel2 0.0000000

plot_kernel(param = param, mix = mix)

plot(mod, type = "kernel")