rankTransPheno {FRGEpistasis} | R Documentation |
Rank-Based Inverse Normal Transformation of Phenotype
rankTransPheno(pheno, para_c)
pheno |
Vector of phenotype which is the quantitative trait. |
para_c |
Adjust parameter, commonly as 0,1/3,3/8 or 1/2. |
Some variables are not normally distributed. And using statistical tests on this data can give misleading results because they do not meet the statistical assumptions. This function implements Rank-Based Inverse Normal Transformation to make phenotype normally distributed.
Return vector of rank-based inverse normal transformed phenotype.
Futao Zhang
T. Mark Beasley, Stephen Erickson and David B. Allison. Rank-Based Inverse Normal Transformations are Increasingly Used, But are They Merited? Behav Genet. 2009 Sep.;39(5):580-595.
c=0.5 smp_num=100 pheno<-sample(c(0:500),smp_num,replace=TRUE) rankTransPheno(pheno,c)