SJIVEfit: IV Estimation and Inference

In vmoprojs/ManyIV: IV Estimation and Inference

Description

It will compute the SJIVE/SJEF estimator and the associated variance covariance matrix

Usage

SJIVEfit(y=NULL,X1=NULL,X2=NULL,Z1=NULL,Z2=NULL,full=2)

Arguments

y	Blah. The default is NULL
X1	Blah. The default is NULL
X2	Blah. The default is NULL
Z1	Blah. The default is NULL
Z2	Blah. The default is NULL
full	Blah. The default is NULL. See Details.

Details

The optimization method is Blah.

The full parameter allows to...

The Z1 parameter allows to...

Value

Returns an object of class IVEstim. An object of class IVEstim is a list containing at most the following components:

param

The vector of parameters' estimates;

Author(s)

Federico Crudu, [e-mail],https://sites.google.com/site/federicocrudu/, Víctor Morales Oñate, [e-mail], https://sites.google.com/site/moralesonatevictor/

References

P. A. Bekker & F. Crudu (2015). "Jackknife instrumental variable estimation with heteroskedasticity",Journal of Econometrics, 185, 332–342.

Examples

library(ManyIV)
# ST: Data Simulation:
y<-rnorm(500)#log(packs)  # dependent variable
X1<- matrix(rnorm(500),500,1)#cbind(log(rprice),log(rincome))# matrix of endogenous variables  
Z1<- matrix(rnorm(10000),500,20)#cbind(log(rincome) , tdiff , I(tax/cpi))# non included instruments
Z2<- matrix(1,length(y),1)# included instruments
# END: Data Simulation:


# ST: Estimation
sol = SJIVEfit(y,X1,X2,Z1,Z2,full=2)
# END: Estimation
str(sol)
plot(density(sol$beta))

vmoprojs/ManyIV documentation built on Feb. 16, 2020, 12:24 a.m.