Follow-up from "Adding dnest4 Sampler"
The following discussions from !849 (merged) should be addressed:
Why are you using estimates of the prior widths and centres? You can get these using the
.maximumattributes of the prior classes.
This line seems superfluous to me, why not just pass
These inline comments seem unnecessary
These print statements should be removed.
All this commented code should be removed.
It looks like this is duplicating a bunch of code from the
pymultinestimplementation. We should aim to reduce this duplication.
FakeSampler, the other samplers are listed in alphabetical order. I know this is minor, but it would be helpful readability,
We don't support python 2 anymore so please don't add
We usually don't use simple
assertstatements, because they are hard to debug for users. Could you add a helpful exception.
if b <= a: raise ValueError("Some helpful information")
Agree with Colm here.
for key in self.search_parameter_keys: width = self.priors[key].maximum - self.priors[key].minimum center = (self.priors[key].maximum + self.priors[key].minimum) / 2.0 widths.append(width) centers.append(center)
The return is unnecessary here