How much does each fold still leak? — fold_leakage

Reports, per fold, the same quantities the pipeline check reports for the fixed split: the share of validation points sharing a raster cell with a training point, and sharing patch pixels at each window. Deliberately the SAME criterion, so a number here can be compared with a number there rather than being a second, incompatible notion of leakage.

Usage

fold_leakage_report(plan, meta, cell_size, windows = c(3L, 9L, 15L))

Arguments

plan: A fold_plan.
meta: Patch store meta with x/y.
cell_size: Raster resolution, in the units of x/y.
windows: Window sizes to report overlap for.

Value

A tibble with a fold column: the overlap shares per role and window, fold by fold.

Examples

ex <- example_landscape()
meta <- data.frame(sample_id = seq_len(nrow(ex$profiles)), x = ex$profiles$x,
                   y = ex$profiles$y)
blocks   <- spatial_folds(meta, k = 3, block_size = 0.05)
buffered <- spatial_folds(meta, k = 3, block_size = 0.05, buffer = 7 * 0.0025)
fold_leakage_report(blocks, meta, cell_size = 0.0025, windows = c(3, 7))
#> # A tibble: 9 × 8
#>    fold split      criterion                  window matters     n   pct n_split
#>   <int> <chr>      <chr>                       <dbl> <lgl>   <int> <dbl>   <int>
#> 1     1 validation identical patch (same ras…     NA TRUE        0  0         55
#> 2     1 validation shares pixels (3x3)             3 FALSE       1  1.82      55
#> 3     1 validation shares pixels (7x7)             7 FALSE      22 40         55
#> 4     2 validation identical patch (same ras…     NA TRUE        0  0         53
#> 5     2 validation shares pixels (3x3)             3 FALSE      15 28.3       53
#> 6     2 validation shares pixels (7x7)             7 FALSE      32 60.4       53
#> 7     3 validation identical patch (same ras…     NA TRUE        0  0         52
#> 8     3 validation shares pixels (3x3)             3 FALSE      12 23.1       52
#> 9     3 validation shares pixels (7x7)             7 FALSE      20 38.5       52
fold_leakage_report(buffered, meta, cell_size = 0.0025, windows = c(3, 7))
#> # A tibble: 9 × 8
#>    fold split      criterion                  window matters     n   pct n_split
#>   <int> <chr>      <chr>                       <dbl> <lgl>   <int> <dbl>   <int>
#> 1     1 validation identical patch (same ras…     NA TRUE        0     0      55
#> 2     1 validation shares pixels (3x3)             3 FALSE       0     0      55
#> 3     1 validation shares pixels (7x7)             7 FALSE       0     0      55
#> 4     2 validation identical patch (same ras…     NA TRUE        0     0      53
#> 5     2 validation shares pixels (3x3)             3 FALSE       0     0      53
#> 6     2 validation shares pixels (7x7)             7 FALSE       0     0      53
#> 7     3 validation identical patch (same ras…     NA TRUE        0     0      52
#> 8     3 validation shares pixels (3x3)             3 FALSE       0     0      52
#> 9     3 validation shares pixels (7x7)             7 FALSE       0     0      52