statsmodels.genmod.generalized_estimating_equations.GEEResults.resid_split static GEEResults.resid_split() [source] Returns the residuals, the endogeneous data minus the fitted values from the model. The residuals are returned as a list of arrays containing the residuals for each cluster.

statsmodels.sandbox.distributions.extras.SkewNorm_gen.entropy SkewNorm_gen.entropy(*args, **kwds) Differential entropy of the RV. Parameters: arg1, arg2, arg3,... : array_like The shape parameter(s) for the distribution (see docstring of the instance object for more information). loc : array_like, optional Location parameter (default=0). scale : array_like, optional Scale parameter (default=1).

statsmodels.miscmodels.count.PoissonZiGMLE.jac PoissonZiGMLE.jac(*args, **kwds) jac is deprecated, use score_obs instead! Use score_obs method. jac will be removed in 0.7. Jacobian/Gradient of log-likelihood evaluated at params for each observation.

statsmodels.genmod.generalized_estimating_equations.GEE.update_cached_means GEE.update_cached_means(mean_params) [source] cached_means should always contain the most recent calculation of the group-wise mean vectors. This function should be called every time the regression parameters are changed, to keep the cached means up to date.

statsmodels.tsa.arima_process.lpol2index statsmodels.tsa.arima_process.lpol2index(ar) [source] remove zeros from lagpolynomial, squeezed representation with index Parameters: ar : array_like coefficients of lag polynomial Returns: coeffs : array non-zero coefficients of lag polynomial index : array index (lags) of lagpolynomial with non-zero elements

statsmodels.sandbox.distributions.extras.SkewNorm_gen.nnlf SkewNorm_gen.nnlf(theta, x) Return negative loglikelihood function Notes This is -sum(log pdf(x, theta), axis=0) where theta are the parameters (including loc and scale).

statsmodels.discrete.discrete_model.MultinomialModel.fit MultinomialModel.fit(start_params=None, method='newton', maxiter=35, full_output=1, disp=1, callback=None, **kwargs) [source] Fit the model using maximum likelihood. The rest of the docstring is from statsmodels.base.model.LikelihoodModel.fit Fit method for likelihood based models Parameters: start_params : array-like, optional Initial guess of the solution for the loglikelihood maximization. The default is an array of zeros. method

statsmodels.stats.weightstats.CompareMeans.ttest_ind CompareMeans.ttest_ind(alternative='two-sided', usevar='pooled', value=0) [source] ttest for the null hypothesis of identical means this should also be the same as onewaygls, except for ddof differences Parameters: x1, x2 : array_like, 1-D or 2-D two independent samples, see notes for 2-D case alternative : string The alternative hypothesis, H1, has to be one of the following ?two-sided?: H1: difference in means not equal to value (def

statsmodels.emplike.descriptive.DescStatUV.test_kurt DescStatUV.test_kurt(kurt0, return_weights=False) [source] Returns -2 x log-likelihood and the p-value for the hypothesized kurtosis. Parameters: kurt0 : float Kurtosis value to be tested return_weights : bool If True, function also returns the weights that maximize the likelihood ratio. Default is False. Returns: test_results : tuple The log-likelihood ratio and p-value of kurt0

statsmodels.tsa.arima_model.ARIMAResults.maroots static ARIMAResults.maroots()