请教关于概率积分变换的问题，您会做了吗？我现在纠结于这个问题，望解答。多谢啊

marginal distributions estimated in Matlab

use function  filtReturnsGARCH.m
  [residuals, UnResiduals,GARCHspec,likelihoods]=filtReturnsGARCH(returns,distr,model,method)

%Estimates standardize residuals and transforms them to uniform
%OUTPUTS:
%residuals:             Matrix TxN with the standardized residuals
%UnResiduals:           Matrix TxN with uniform marginals (input for copula)

method 'CML': use empiricalCDF()  % transform margin to uniform
                  UnResiduals(:,i) =empiricalCDF(residuals(:,i));  

method 'IFM': if you knows the distribution of the residuals
               distr,'Gaussian'   UnResiduals(:,i) =normcdf(residuals(:,i));
               distr,'T'   UnResiduals(:,i) =tcdf(residuals(:,i),...);
               distr,'SkewT'   UnResiduals(:,i) =skewtdis_cdf(residuals(:,i),.....);

这种可以使用经验cdf来转换吧

X<-as.matrix(UKlogreturns)
Y<-as.matrix(EUROlogreturns)
XY<-cbind(X,Y)
XY<-XY[XY[,1]!=0&XY[,2]!=0,]
XY<-na.omit(XY)
copulaXY<-apply(XY,2,edf,adjust=1)

用apply函数做，参数是“edf"

library(fGarch)
m1=garchFit(~garch(1,1),data=rsh,trace=F) #garch(1,1)
res1=residuals(m1,standardize=TRUE) #提取标准化残差
pres1=pnorm(res1)#概率积分转换
ks.test(pres1,"punif")
我运行出来p值很小，程序有错吗？