Large increase in precision for the CaM frame order isotropic cone model test data set.
The higher precision is because the number structures in the distribution is now 20 million rather
than 1 million and the numpy.float128 data averaging has been used.