<< olsc0 Single equation regressions olspec >>

Grocer >> Single equation regressions > olsc1

olsc1

Cochrane-Orcutt estimation of an autocorrelated model

CALLING SEQUENCE

rolsc=olsc1(y,x,maxit,crit)

PARAMETERS

Input

* y = real (nx1) vector of an endogenous variable

* x = real (nxk) vector of an exogenous variable

* maxit = a scalar, the maximum # of authorized iterations

* crit = a scalar, the convergence criterionused to assess if the difference between successive values of the autocorrelation coefficient is significant

 

Output

* rolsc = a results tlist with

  - rolsc('meth') = 'Cochrane-Orcutt'

  - rolsc('y') = y data vector

  - rolsc('x') = x data matrix

  - rolsc('nobs') = # observations

  - rolsc('nvar') = # variables

  - rolsc('beta') = bhat

  - rolsc('yhat') = yhat

  - rolsc('resid') = residuals

  - rolsc('vcovar') = estimated variance-covariance matrix of beta

  - rolsc('sige') = estimated variance of the residuals

  - rolsc('sigu') = sum of squared residuals

  - rolsc('ser') = standard error of the regression

  - rolsc('tstat') = t-stats

  - rolsc('pvalue') = pvalue of the betas

  - rolsc('dw') = Durbin-Watson Statistic

  - rolsc('condindex') = multicolinearity cond index

  - rolsc('prescte') = boolean indicating the presence or absence of a constant in the regression

  - rolsc('rsqr') = rsquared

  - rolsc('rbar') = rbar-squared

  - rolsc('f') = F-stat for the nullity of coefficients other than the constant

  - rolsc('pvaluef') = its significance level

  - rolsc('bounds') = if there is a timeseries in the regression, the bounds of the regression

  - rolsc('rho') = estimated first order autocorrelation of residuals

  - rolsc('trho') = its Student t

  - rolsc('iterout') = a (niter x 3) matrix giving for each

DESCRIPTION

Computes Cochrane-Orcutt ols regression for AR1 errors. This is a more constrainted version than olsc: endogenous variables and exogenous ones are given in a matrix form; the maximum number of iterations and the criterion must be given. No result is displayed on screen.

EXAMPLE

x=grand(15,3,'nor',0,1)
y=x*ones(3,1)+autocum(grand(15,1,'nor',0,1),0.8)
rolsc=olsc1(y,x,100,%eps^(1/3))
//In this example, the maximum # of iterations is set to 100 and the convergence criterion to %eps^(1/3).

AUTHOR

Eric Dubois 2005

Report an issue
<< olsc0 Single equation regressions olspec >>