6.3.3.2.2.3. statsmodels.sandbox.distributions.estimators.fitbinnedgmm¶

statsmodels.sandbox.distributions.estimators.fitbinnedgmm(distfn, freq, binedges, start, fixed=None, weightsoptimal=True)[source]¶

estimate parameters of distribution function for binned data using GMM

Parameters:

distfn : distribution instance

needs to have cdf method, as in scipy.stats

freq : array, 1d

frequency count, e.g. obtained by histogram

binedges : array, 1d

binedges including lower and upper bound

start : tuple or array_like ?

starting values, needs to have correct length

fixed : None

not used yet

weightsoptimal : boolean

If true, then the optimal weighting matrix for GMM is used. If false, then the identity matrix is used

Returns:

paramest : array

estimated parameters

Notes

todo: add fixed parameter option

added factorial