statsmodels.tools.eval_measures.aicc¶
-
statsmodels.tools.eval_measures.
aicc
(llf, nobs, df_modelwc)[source]¶ Akaike information criterion (AIC) with small sample correction
- Parameters
- llffloat
value of the loglikelihood
- nobsint
number of observations
- df_modelwcint
number of parameters including constant
- Returns
- aiccfloat
information criterion
References
http://en.wikipedia.org/wiki/Akaike_information_criterion#AICc