##
## Wilcoxon signed rank test with continuity correction
##
## data: as.vector(lvl4_acc_ct) and as.vector(acc_leave1out_ct)
## V = 5065, p-value = 0.0001294
## alternative hypothesis: true location shift is not equal to 0
## 95 percent confidence interval:
## -0.07124282 -0.02337594
## sample estimates:
## (pseudo)median
## -0.04759526
## [1] "Mean accuracy of control: 32%"
## [1] "Mean accuracy of withheld: 28%"
## [1] "Median pairwise change in accuracy in withheld cell lines relative to control was -5% (p = 0.00013 by Wilcoxon signed-rank test)."
##
## Wilcoxon signed rank test with continuity correction
##
## data: temp_acc$lvl5L1O and temp_acc$lvl4CT
## V = 10160, p-value = 0.0001294
## alternative hypothesis: true location shift is not equal to 0
Redone with multiclass MCC score (from Yardstick package) instead of accuracy.
## [1] "Lvl5 leave one out (CTRL) MCC: 0.331336580374155"