In #37 I opened an issue that the example gaussian.py does not produce a sampled log_Z which is close to the analytical log_Z. In order to further debug this, I added the same example to the C++ version where I get:
When checking the posterior samples everything looks perfect also:
So the C++-version seems to work. I will now move on to explicitly check the posteriors for the python example.
I mainly added this so that my code is visible and maybe it will help someone else in the future.
In #37 I opened an issue that the example gaussian.py does not produce a sampled log_Z which is close to the analytical log_Z. In order to further debug this, I added the same example to the C++ version where I get:
When checking the posterior samples everything looks perfect also:
So the C++-version seems to work. I will now move on to explicitly check the posteriors for the python example. I mainly added this so that my code is visible and maybe it will help someone else in the future.