From Jay B: "I found that the sample size for estimating bootstrap values of ESW was much smaller than the total sample size. I realized that this was because the bootstrap sample (sampling with replacement) will necessarily have duplicates of many sightings. However your LTabundR::lta function eliminates duplicates when estimating ESW. This makes sense because you only want to count each sighting once (even though it may contain multiple species on separate lines). However, it doesn't make sense in the context of a bootstrap."
Resolve this within lta() and ensure this is not also happening within lta_subgroup().
Not an issue after all, LTabundR already handled the duplicated sighting numbers introduced by bootstrapping, and all sample sizes are as they should be. Closing.
From Jay B: "I found that the sample size for estimating bootstrap values of ESW was much smaller than the total sample size. I realized that this was because the bootstrap sample (sampling with replacement) will necessarily have duplicates of many sightings. However your LTabundR::lta function eliminates duplicates when estimating ESW. This makes sense because you only want to count each sighting once (even though it may contain multiple species on separate lines). However, it doesn't make sense in the context of a bootstrap."
Resolve this within
lta()
and ensure this is not also happening withinlta_subgroup()
.