There’s a typo in the book: the numerator of the ”f” function needs to be squared to recover Le Cam’s divergence ;)