cuqi.distribution.Normal.enable_FD#
- Normal.enable_FD(epsilon=1e-08)#
Enable finite difference approximation for logd gradient. Note that if enabled, the FD approximation will be used even if the _gradient method is implemented.
- Parameters:
epsilon (float)
logd (Spacing (step size) to use for finite difference approximation for)
1e-8. (gradient for each variable. Default is)