Possible optimization of decision rules at first order
In dr.pdf, at the end of section 4.1, we have:
g^-_y = Z'11 T^{-1}11 S{11} (Z'{11})^{-1}
But dr1.m used to contain the following formula:
g^-_y = (T^{-1}11 X)^{-1} S{11} X
which is probably more efficient given that X=Z_{11}+Z_{12}g⁺_y (see equation 10), because it saves an inversion
We should make benchmarks between the 2 formulas, because Ondra thinks that for performance reasons, the first method is better, even though this is not obvious.