statsmodels.sandbox.regression.gmm.IV2SLS

class statsmodels.sandbox.regression.gmm.IV2SLS(endog, exog, instrument=None)[source]

Instrumental variables estimation using Two-Stage Least-Squares (2SLS)

Parameters:

endog: array

Endogenous variable, 1-dimensional or 2-dimensional array nobs by 1

exog : array

Explanatory variables, 1-dimensional or 2-dimensional array nobs by k

instruments : array

Instruments for explanatory variables. Must contain both exog variables that are not being instrumented and instruments

Notes

All variables in exog are instrumented in the calculations. If variables in exog are not supposed to be instrumented, then these variables must also to be included in the instrument array.

Degrees of freedom in the calculation of the standard errors uses df_resid = (nobs - k_vars). (This corresponds to the small option in Stata’s ivreg2.)

Attributes

endog_names Names of endogenous variables
exog_names Names of exogenous variables

Methods

fit() estimate model using 2SLS IV regression
from_formula(formula, data[, subset, drop_cols]) Create a Model from a formula and dataframe.
hessian(params) The Hessian matrix of the model
information(params) Fisher information matrix of model
initialize()
loglike(params) Log-likelihood of model.
predict(params[, exog]) Return linear predicted values from a design matrix.
score(params) Score vector of model.
whiten(X)

Methods

fit() estimate model using 2SLS IV regression
from_formula(formula, data[, subset, drop_cols]) Create a Model from a formula and dataframe.
hessian(params) The Hessian matrix of the model
information(params) Fisher information matrix of model
initialize()
loglike(params) Log-likelihood of model.
predict(params[, exog]) Return linear predicted values from a design matrix.
score(params) Score vector of model.
whiten(X)

Attributes

endog_names Names of endogenous variables
exog_names Names of exogenous variables