Skip to content

Instantly share code, notes, and snippets.

@sTeamTraen
sTeamTraen / MaghbooliFig1.csv
Last active September 29, 2020 22:11
Reverse-engineered data for Maghbooli et al. (10.1371/journal.pone.0239799) Figure 1. Extracted by Bob Sterman. Age, alive/dead status, and VitD levels above 30.0 confirmed by Nick Brown.
Num X Y Age VitD Status
1 248 767 37 107.59 A
2 552 830 60 118.45 A
3 870 524 84 65.69 A
4 777 511 77 63.45 A
5 301 471 41 56.55 A
6 420 464 50 55.34 A
7 486 513 55 63.79 A
8 579 502 62 61.90 A
9 632 484 66 58.79 A
@sTeamTraen
sTeamTraen / Effect-size-D.R
Last active October 14, 2019 11:34
An interesting phenomenon with (perhaps extreme?) data in PET-PEESE
set.seed(1)
# Number of datasets to generate.
iter <- 1000
# Generate random sample sizes.
# Initially I only put the Nmin parameter there to avoid silly sizes like 0 or 1,
# but then I found that something interesting happens.
# Try setting Nmean to 10 instead of 50, so that all sample sizes are 20, and watch
# what happens to the correlation between d and its SE.