Sub-sample a spectra collection

Pick a subsample form a spectra collection

subsample(x, ...)

# S3 method for collection
subsample(x, size = 2, balance_labels = FALSE, ids = NULL, ...)

Arguments

x	Acollection object
...	further arguments passed to or from other methods(not currenctly used).
size	Size of the sub-sampling
balance_labels	A boolean, wether to pick spectra equally between labels. If the subsampling size is below the number of labels, not all labels will be chosen from.
ids	A vector of ids to extract, this will override the other parameters.

Value

a subsample of x

Examples

library(tidySpectR)
library(dplyr)

# Randomly sample
subsample(fa_nmr, size = 3)
#> Spectra collection containing 3 entries.
#> Number of bins: 2250 
#> Limits: -2.123166e-05 8.999979 
#> Labels: conventional organic 
#> 
#> Processing:

# Sample one spectra for each labels
subsample(fa_nmr, size= 2, balance_labels = TRUE)
#> Spectra collection containing 2 entries.
#> Number of bins: 2250 
#> Limits: -2.123166e-05 8.999979 
#> Labels: conventional organic 
#> 
#> Processing:

# Subsample specific IDs
subsample(fa_nmr, ids = c("20199305928", "20199306281", "20199504645"))
#> Spectra collection containing 3 entries.
#> Number of bins: 2250 
#> Limits: -2.123166e-05 8.999979 
#> Labels: conventional organic 
#> 
#> Processing:

# Sample alll spectra of specific labels
organic <- fa_nmr$labels %>% 
           filter(label == 'organic') %>%
           pull(id)
subsample(fa_nmr, ids = organic)
#> Spectra collection containing 3 entries.
#> Number of bins: 2250 
#> Limits: -2.123166e-05 8.999979 
#> Labels: conventional organic 
#> 
#> Processing:

# Chain subsamplings to randomly select a spectra with a specific label
subsample(fa_nmr, ids = organic) %>%
  subsample(size=1)
#> Spectra collection containing 1 entries.
#> Number of bins: 2250 
#> Limits: -2.123166e-05 8.999979 
#> Labels: conventional organic 
#> 
#> Processing: