search

kassambara / rstatix

Pipe-friendly Framework for Basic Statistical Tests in R
https://rpkgs.datanovia.com/rstatix/
432 stars 51 forks source link

strange difference between p.adjust and adjust_pvalue methods #193

Open lmanchon opened 10 months ago

lmanchon commented 10 months ago

--Hi,

i don't understand why the 2 methods give different results:

set.seed(123) x <- rnorm(50, mean = c(rep(0, 25), rep(3, 25))) p <- 2*pnorm(sort(-abs(x)))

p.adjust(p, "BH") [1] 1.177027e-05 4.294736e-04 4.294736e-04 9.471343e-04 9.471343e-04 9.471343e-04 9.471343e-04 9.803146e-04 1.252532e-03 1.897690e-03 2.779520e-03 6.724807e-03 1.263693e-02 1.263693e-02 [15] 1.745666e-02 2.076243e-02 2.076243e-02 2.445869e-02 2.473958e-02 2.824495e-02 5.035754e-02 1.118804e-01 1.304633e-01 1.304633e-01 1.479031e-01 1.592520e-01 1.598765e-01 2.126168e-01 [29] 3.259994e-01 3.430823e-01 3.563248e-01 4.462501e-01 4.619538e-01 6.835770e-01 6.835770e-01 6.835770e-01 7.188450e-01 7.414352e-01 7.414352e-01 7.626063e-01 7.626063e-01 7.626063e-01 [43] 7.626063e-01 7.824868e-01 7.988737e-01 8.802645e-01 8.802645e-01 9.304775e-01 9.304775e-01 9.437890e-01

p %>% adjust_pvalue(method = "BH") [1] 2.354054e-07 2.101590e-05 2.576842e-05 9.814783e-05 1.052610e-04 1.241481e-04 1.325988e-04 1.568503e-04 2.254557e-04 3.795380e-04 6.114943e-04 1.613954e-03 3.302430e-03 3.538342e-03 [15] 5.236997e-03 6.831909e-03 7.059226e-03 8.805129e-03 9.401040e-03 1.129798e-02 2.115017e-02 4.922736e-02 6.053298e-02 6.262239e-02 7.395153e-02 8.281103e-02 8.633331e-02 1.190654e-01 [29] 1.890796e-01 2.058494e-01 2.209214e-01 2.856000e-01 3.048895e-01 4.660682e-01 4.830809e-01 4.921755e-01 5.319453e-01 5.751550e-01 5.783195e-01 6.185894e-01 6.363620e-01 6.448587e-01 [43] 6.558414e-01 6.885884e-01 7.189864e-01 8.179539e-01 8.274487e-01 8.971300e-01 9.118680e-01 9.437890e-01

lmanchon commented 10 months ago

problem is over sorry. I forgot to detach stats package and rstatix was masked by stats.