Sample size determination for the association between longitudinal and time‐to‐event outcomes using the joint modeling time‐dependent slopes parameterization