<< Qualitative Econometrics Functions Qualitative Econometrics Functions multilogit >>

Grocer >> Qualitative Econometrics Functions > logit

logit

logit regression

CALLING SEQUENCE

[rlogit]=logit(namey,arg1,...,argn)

PARAMETERS

Input

* namey = a time series, a real (nx1) vector or a string equal to the name of a time series or a (nx1) real vector between quotes

* argi = an argument which can be:

  - a time series

  - a real (nx1) vector

  - a real (nxk) matrix

  - a string equal to the name of a time series or a (nxk) real vector or matrix between quotes

  - a list of such elements

  - the string 'noprint' if the user doesn't want to display the results of the regression

  - the string 'maxit=xx' if the user wants to set the maximum # of iterations to xx (default=100)

  - the string 'tol=xx' if the user wants to set the convergence criterion to xx (default=1e-6)

 

Output

* rlogit = a results tlist with

  - rlogit('meth') = 'logit'

  - rlogit('y') = y data vector

  - rlogit('x') = x data matrix

  - rlogit('nobs') = # observations

  - rlogit('nvar') = # variables

  - rlogit('beta') = bhat

  - rlogit('yhat') = yhat

  - rlogit('resid') = residuals

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

  - rlogit('tstat') = t-stats

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

  - rlogit('r2mf') = = McFadden pseudo-R²

  - rlogit('rsqr') = = Estrella R²

  - rlogit('lratio') = LR-ratio test against intercept model

  - rlogit('lik') = unrestricted Likelihood

  - rlogit('zip') = # of 0's

  - rlogit('one) = # of 1's

  - rlogit('iter') = # of iterations

  - rlogit('crit') = convergence criterion

  - rlogit('namey') = name of the y variable

  - rlogit('namex') = name of the x variables

  - rlogit('prests') = boolean indicating the presence or absence of a time series in the regression

  - rlogit('prescte') = %f (for printings)

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

DESCRIPTION

Computes Logit Regression. If the user has not given the argument 'noprint', displays on screen the results of the regression and various diagnostics. References: Arturo Estrella (1998) 'A new measure of fit for equations with dichotomous dependent variable', JBES, Vol. 16, #2, April, 1998.

EXAMPLE

grade = zeros(32,1);
grade([5 10 14 20 22 25:27 29 30 32],1)  = 1;
gra=reshape(grade,'1a')
 
// psi variable
psi = reshape([zeros(18,1) ; ones(14,1)],'1a')
tuce = reshape([20 22 24 12 21 17 17 21 25 29 20 23 23 25 26 19 ...
        25 19 23 25 22 28 14 26 24 27 17 24 21 23 21 19]','1a')
 
gpa = reshape([2.66 2.89 3.28 2.92 4.00 2.86 2.76 2.87 3.03 3.92 ...
       2.63 3.32 3.57 3.26 3.53 2.74 2.75 2.83 3.12 3.16 ...
       2.06 3.62 2.89 3.51 3.54 2.83 3.39 2.67 3.65 4.00 ...
       3.10 2.39]','1a')
 
bounds()
logit('grade','cte','psi','tuce','gpa');
// Example, taken from logit_d, provides the logit regression of vector grade on a constant and the vectors psi, tuce, gpa.
 
logit('grade','cte','psi','tuce','gpa','maxit=200','crit=sqrt(%eps)','noprint');
// Does the same, except that the maximum number of iterations is set to 200 (instead of 100),
// the convergence criterion to sqrt(%eps) (instead of 0.000001)
// and the results are not displayed on screen.

AUTHOR

Eric Dubois 2002

Report an issue
<< Qualitative Econometrics Functions Qualitative Econometrics Functions multilogit >>