| PGC {GenABEL} | R Documentation |
This function estimates the genomic controls for different models and degrees of freedom, using polinomial function. Polinomial coefficients are estimated by optimizing different error functions: regress, median, ks.test or group regress.
PGC(data, method = "group_regress", p, df, pol.d = 3,
plot = TRUE, index.filter = NULL, start.corr = FALSE,
proportion = 1, n_quiantile = 5, title_name = "Lambda",
type_of_plot = "plot", lmax = NULL, color = "red")
data |
Input vector of Chi square statistic |
method |
Function of error to be optimized. Can be "regress", "median", "ks.test" or "group_regress" |
p |
Input vector of allele frequencies |
df |
Number of degrees of freedom |
pol.d |
The degree of polinomial function |
plot |
If TRUE, plot of lambda will be produced |
start.corr |
For regress method use it only when you want to make calculations faster |
index.filter |
Index of variables in data vector, that will be used in analysis if zero - all variables will be used |
proportion |
The proportion of lowest P (Chi2) to be used when estimating the inflation factor Lambda for "regress" method only |
n_quiantile |
The number of groups for "group_regress" method |
title_name |
The title name for plot |
type_of_plot |
For developers only |
lmax |
The threshold for lambda for plotting (optional) |
color |
The color of the plot |
A list with elements
data |
Output vector corrected Chi square statistic |
b |
Polinomial coefficients |
Yakov Tsepilov
require(GenABEL.data) data(ge03d2) ge03d2 <- ge03d2[seq(from=1,to=nids(ge03d2),by=2),seq(from=1,to=nsnps(ge03d2),by=3)] qts <- mlreg(dm2~1,data=ge03d2,gtmode = "additive") chi2.1df <- results(qts)$chi2.1df s <- summary(ge03d2) freq <- s$Q.2 result=PGC(data=chi2.1df,method="median",p=freq,df=1, pol.d=2, plot=TRUE, lmax=1.1,start.corr=FALSE)