<< olst Single equation regressions olstmod >>
grocer >> Single equation regressions > olst1

olst1

ols with t distributed errors

CALLING SEQUENCE

[rolst]=olst1(y,x,crit,maxit)

PARAMETERS

Input

* y = a (n x 1) vector

* x = a (n x k) vector

* crit = a scalar, the convergence criterion (for example 1e-5' ; default = 1e-15)

* maxit = a scalar, the maximum # of iterations (for example 'itmax=10'; default=500)

Output

* rolst = a tlist with

   - rolst('meth') = 'olst'

   - rolst('y') = y data vector

   - rolst('x') = x data matrix

   - rolst('nobs') = nobs

   - rolst('nvar') = nvars

   - rolst('beta') = bhat

   - rolst('yhat') = yhat

   - rolst('resid') = residuals

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

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

   - rolst('sigu') = estimated sum of squared residuals

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

   - rolst('tstat') = t-stats

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

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

   - rolst('condindex') = multicolinearity cond index

   - rolst('conv') = convergence max(abs(bnew-bold))

   - rolst('iter') = # of iterations

   - rolst('prescte') = boolean indicating if the Rª can be calculated

DESCRIPTION

Peforms ols with t-distributed errors. The low-level counrerpart to oslt, working only with matrices.

EXAMPLE

b = ones(5,1);
xmat = rand(100,4,'n');
exo = [ones(100,1),xmat];
// generate t-distributed errors
 
evec = tdis_rnd(100,2)*.5;
endo = exo*b+evec;
// do robust t-distributed errors regression:
lresult = olst1(endo,exo,0.0001,1000);

AUTHOR

Éric Dubois 2019

Report an issue
<< olst Single equation regressions olstmod >>