Suppose we are fitting a linear regression model with n samples and pp predictors (including the intercept). Circle correct statements.

There are three correct statements.

-The LS coefficients are not unique if a non-intercept predictor takes a constant value

-The LS coefficients are not unique if p is bigger than n.

-The LS coefficients are unique only if the predictors are uncorrelated with each other.

-The LS coefficients are not unique if columns of the n-by-p design matrix \mathbf{X}X are linear dependent.

-The LS coefficients are not unique if the response variable Y takes a constant value