> Non-obvious behaviors
gotchas
Explore
115,249 skills indexed with the new KISS metadata standard.
gotchas
`$p.value`
not of the distribution itself.
`shape2` (beta) — no `mean`/`sd` parameterization.
linear interpolation). Type 1 = inverse of empirical CDF (SAS default). Types 4-9 are continuous; 1-3 are discontinuous.
uses `length(n)` as the count
adding 0%
N-1 vs N scaling).
`maximum`
`BH` (Benjamini-Hochberg FDR)
uses simulation (`simulate.p.value = TRUE`) or network algorithm.
not rank-sum (Mann-Whitney).
normal approximation used.
`$residuals` (Pearson)
not the chi-sq assumption).
not a matrix. `p` must sum to 1 (or set `rescale.p = TRUE`).
gotchas
no other stopping criterion.
upper]`. **Requires** `f(lower)` and `f(upper)` to have opposite signs.
model2)`: F-test comparing nested models.
`print`
robust but slow). Poor for 1D — use `Brent` or `optimize()`.
default 2/3); `loess` parameter is `span` (default 0.75).
`forward`