Fit classification tree to klassets_response_xy object

Fit classification tree to klassets_response_xy object

Usage

fit_classification_tree(df, type = "prob", maxdepth = Inf, alpha = 0.05, ...)

Arguments

df

A object from sim_response_xy.

type

Type of prediction, one of prob, response, node.

maxdepth

Max depth of the tree. Same used in partykit::ctree_control.

alpha

Alpha value, same used in partykit::ctree_control

...

Options for partykit::ctree_control.

Examples

set.seed(123)

df <- sim_response_xy(n = 1000, relationship = function(x, y) x**2 > sin(y))

plot(df)


# default type = "prob"
df_tree_prob <- fit_classification_tree(df)
df_tree_prob
#> # A tibble: 1,000 × 4
#>    response       x       y prediction
#>    <fct>      <dbl>   <dbl>      <dbl>
#>  1 FALSE    -0.425  -0.453       0.242
#>  2 FALSE     0.577   0.188       0.564
#>  3 FALSE    -0.182  -0.680       0.242
#>  4 FALSE     0.766   0.707       0.746
#>  5 TRUE      0.881   0.695       0.746
#>  6 TRUE     -0.909  -0.0442      0.242
#>  7 FALSE     0.0562  0.547       0.564
#>  8 FALSE     0.785  -0.409       0.242
#>  9 FALSE     0.103  -0.869       0.242
#> 10 TRUE     -0.0868 -0.119       0.242
#> # … with 990 more rows

df_tree_resp <- fit_classification_tree(df, type = "response")
df_tree_resp
#> # A tibble: 1,000 × 4
#>    response       x       y prediction
#>    <fct>      <dbl>   <dbl> <fct>     
#>  1 FALSE    -0.425  -0.453  TRUE      
#>  2 FALSE     0.577   0.188  FALSE     
#>  3 FALSE    -0.182  -0.680  TRUE      
#>  4 FALSE     0.766   0.707  FALSE     
#>  5 TRUE      0.881   0.695  FALSE     
#>  6 TRUE     -0.909  -0.0442 TRUE      
#>  7 FALSE     0.0562  0.547  FALSE     
#>  8 FALSE     0.785  -0.409  TRUE      
#>  9 FALSE     0.103  -0.869  TRUE      
#> 10 TRUE     -0.0868 -0.119  TRUE      
#> # … with 990 more rows

df_tree_node <- fit_classification_tree(df, type = "node")
df_tree_node
#> # A tibble: 1,000 × 4
#>    response       x       y prediction
#>    <fct>      <dbl>   <dbl>      <int>
#>  1 FALSE    -0.425  -0.453           2
#>  2 FALSE     0.577   0.188           4
#>  3 FALSE    -0.182  -0.680           2
#>  4 FALSE     0.766   0.707           5
#>  5 TRUE      0.881   0.695           5
#>  6 TRUE     -0.909  -0.0442          2
#>  7 FALSE     0.0562  0.547           4
#>  8 FALSE     0.785  -0.409           2
#>  9 FALSE     0.103  -0.869           2
#> 10 TRUE     -0.0868 -0.119           2
#> # … with 990 more rows

plot(df_tree_prob)

plot(df_tree_resp)

plot(df_tree_node)