Extract directional search points during the optimisation

get_dir_search(dt, ratio = 5, ...)

Arguments

dt

a data object collected by the projection pursuit guided tour optimisation in the tourr package

ratio

numeric; a buffer value to deviate directional search points from the anchor points

...

arguments passed to compute_pca()

Value

a tibble object containing the directional search bases in pseudo derivative search

See also

Examples

holes_1d_geo %>% compute_pca() %>% purrr::pluck("aug") %>% get_dir_search()
#> # A tibble: 330 x 16 #> basis index_val info method alpha tries loop id row_num PC1 PC2 #> <list> <dbl> <chr> <chr> <dbl> <dbl> <dbl> <dbl> <int> <dbl> <dbl> #> 1 <dbl[,… 0.749 direct… PD NA 2 1 2 2 -1.80 -0.776 #> 2 <dbl[,… 0.749 direct… PD NA 2 1 3 3 -1.77 -0.601 #> 3 <dbl[,… 0.749 direct… PD NA 2 1 4 4 -1.79 -0.632 #> 4 <dbl[,… 0.749 direct… PD NA 2 1 5 5 -1.78 -0.745 #> 5 <dbl[,… 0.749 direct… PD NA 2 1 6 6 -1.80 -0.562 #> 6 <dbl[,… 0.749 direct… PD NA 2 1 7 7 -1.77 -0.815 #> 7 <dbl[,… 0.749 direct… PD NA 2 1 8 8 -1.73 -0.732 #> 8 <dbl[,… 0.749 direct… PD NA 2 1 9 9 -1.84 -0.644 #> 9 <dbl[,… 0.749 best_d… PD NA 2 1 10 10 -1.78 -0.804 #> 10 <dbl[,… 0.749 direct… PD NA 2 1 11 11 -1.79 -0.573 #> # … with 320 more rows, and 5 more variables: PC3 <dbl>, PC4 <dbl>, PC5 <dbl>, #> # anchor_x <dbl>, anchor_y <dbl>