class IV2SLS ( LikelihoodModel ): ''' Instrumental variables estimation using Two-Stage Least-Squares (2SLS) Parameters - endog : ndarray Endogenous variable, 1-dimensional or 2-dimensional array nobs by 1 exog : ndarray Explanatory variables, 1-dimensional or 2-dimensional array nobs by k instrument : ndarray Instruments for explanatory variables. Author: josef-pktd License: BSD (3-clause) ''' from import lrange import numpy as np from scipy import optimize, stats from import approx_fprime from import ( Model, LikelihoodModel, LikelihoodModelResults ) from _model import ( OLS, RegressionResults, RegressionResultsWrapper ) import _covariance as smcov from import cache_readonly from import _ensure_2d DEBUG = 0 def maxabs ( x ): '''just a shortcut to np.abs(x).max() ''' return np. And I'm confusing between weights and omega, which are *not* the same in this case. I do not have an estimate for omega in that case. jval in jtest looks to large in example, but I have no idea about the size * bse for fitonce look too large (no time for checking now) formula for calc_cov_params for the case without optimal weighting matrix is wrong. Scaling of jval is irrelevant for estimation. Unclear - * dof in Hausman - based on rank - differs between IV2SLS method and function used with GMM or (IV2SLS) - with GMM, covariance matrix difference has negative eigenvalues in iv example, ? * jtest/jval - I'm not sure about the normalization (multiply or divide by nobs) in jtest. What's the best way to do this, multiple inheritance, outsourcing the functions, mixins or delegation (a model creates a GMM instance just for estimation). predict, fitted values, calculating the error term, and some result statistics. It would be good if we can reuse methods from regressions, e.g.

DONE `fitonce` in DistQuantilesGMM, params are the same as in direct call to fitgmm move it to GMM class (once it's clearer for which cases I need this.) * GMM does not know anything about the underlying model, e.g. cov_params, I have it in the standalone function (and in calc_covparams which is a copy of it), but not tested yet. > add result statistics for this case, e.g. iter=1 works (weights is identity matrix) -> need method to do one iteration with an identity matrix or an analytical weighting matrix given as parameter. '''Generalized Method of Moments, GMM, and Two-Stage Least Squares for instrumental variables IV2SLS Issues - * number of parameters, nparams, and starting values for parameters Where to put them? start was initially taken from global scope (bug) * When optimal weighting matrix cannot be calculated numerically In DistQuantilesGMM, we only have one row of moment conditions, not a moment condition for each observation, calculation for cov of moments breaks down.

Stata gmm