resids_diag {GWAS.BAYES} | R Documentation |
Tests residuals to see if they are normal. This looks at the model with all significant SNPs from the preselection phase.
resids_diag(Y,SNPs,significant,kinship = FALSE,principal_components = FALSE,plot_it = TRUE)
Y |
The phenotype response of interest |
SNPs |
Standardized SNP data set where the values of each column are either 0 or 1 |
significant |
A vector of 0's and 1's where the 1's indicate a significant SNP. This is returned in the output of the preselection function. |
kinship |
A kinship matrix, can be calculated from the rrBLUP package. |
principal_components |
A matrix or vector of the principal components one would like to include in the analysis. |
plot_it |
If TRUE a histogram of the residuals is returned. |
value 1 |
The output of a Shapiro-Wilk test for the residuals. If the p-value is above .05, there is no evidence that the residuals are not normal. If the p-value is below .05 there is evidence that the residuals are not normal, and some transformation is suggested. |
value 2 |
A histogram of the residuals when plot_it = TRUE |
data("vignette_lm_dat") Y <- vignette_lm_dat$Phenotype SNPs <- vignette_lm_dat[,-1] fullPreprocess <- preprocess_SNPs(SNPs = SNPs,Y = Y,MAF = 0.01,number_cores = 1) SNPs <- fullPreprocess$SNPs Y <- fullPreprocess$Y fullPreprocess$SNPs_Dropped principal_comp <- pca_function(SNPs = SNPs,number_components = 1,plot_it = FALSE) Significant_SNPs <- preselection(Y = Y, SNPs = SNPs,number_cores = 1, principal_components = principal_comp,frequentist = TRUE,controlrate = "bonferroni",threshold = .05,kinship = FALSE,info = FALSE) resids_diag(Y = Y,SNPs = SNPs,significant = Significant_SNPs$Significant,kinship = FALSE,principal_components = principal_comp,plot_it = TRUE)