Ftest {MLGL} | R Documentation |
Perform a F-test
Ftest(X, y, varToTest)
X |
design matrix of size n*p |
y |
response vector of length n |
varToTest |
vector containing the index of the column of X to test |
y = X * beta + epsilon
null hypothesis : beta[varToTest] = 0 alternative hypothesis : it exists an index k in varToTest such that beta[k] != 0
The test statistic is based on a full and a reduced model. full : y = X * beta[varToTest] + epsilon reduced : the null model
a vector of the same length as varToTest containing the p-values of the test.