Negative binomial fit with Frank Copula
It uses two negative binomial PMFs in the form:
where the mean is m and the variance coupled by the Frank copula in the form:
where u and v are two negative binomial CDFs, to fit a provided numerical probability distribution i.e. a normalized bidimensional histogram. It gives back the two means and r parameters of the marginals, as well as the parameter .
In the main of the py file there is a working example (requires matplotlib).
Requirements:
- Python 3.5 or higher
- Scipy
- Numpy
- MatPlotLib (to visualize the example).