Yashu Seth

@ankurankan Okay :)

@kris-singh Just fixing doc-tests will not suffice. ```Discretize``` methods for ```ContinuousFactor``` has not been implemented.

@kris-singh You need to add tests for the changes you have made.

@kris-singh You have added the functionality to discretize univariate ```ContinuousFactor``` right?

@kris-singh But only for univariate cases. So you will have to add tests for discretize method taking univariate Continuous Factors.

## Reference Probabilistic Graphical Models: Principles and Techniques, Daphne Koller and Nir Friedman, Section 17.2.4 and 17.2.5

@raghavg7796 Please mention the issues you have addressed from #722, using a checkbox or something else. It would be easier to review the changes.

#752

I use windows. I will look into this. On 8 Oct 2016 15:33, "Utkarsh" [email protected] wrote: > @ankurankan https://github.com/ankurankan I don't have a windows > machine. @yashu-seth https://github.com/yashu-seth used to...

@khalibartan I am getting the same error message. I looked into it and found that the value of `np.linalg.norm(sample_covariance - self.test_model.covariance)` comes out to be 0.557192037439 which is not less...