Name

multilogit1 — multivariate logit regression

CALLING SEQUENCE

res = multilogit1(y,x,bet0,maxit,tol)

PARAMETERS

Input

• y = response variable vector (nobs x 1)

 - the response variable should be coded sequentially from 0 to ncat, i.e., y in (0,1,2,…,ncat) entries

• x = matrix of covariates (nobs x nvar)

• bet0 = optional starting values for bet (nvar x ncat+1) (default=0)

• maxit = optional maximum number of iterations (default=100)

• tol = optional convergence tolerance (default=1e-6)

Output

• res = a results tlist with

  - res('meth') = 'multilogit'

  - res('beta') = (nvar x ncat) matrix of bet coefficients: [bet_1 bet_2 … bet_ncat] under the normalization bet_0 = 0

  - res('coeff') = (nvar*ncat x 1) vector of beta coefficients: [beta_1 ; beta_2 ; … ; beta_ncat] under normalization beta_0 = 0

  - res('covb') = (nvar*ncat x nvar*ncat) covariance matrix of coefficients

  - res('tstat') = (nvar*ncat x 1) vector of t-statistics

  - res('pvalue') = (nvar*ncat x 1) vector of corresponding p-values

  - res('y') = (nobs x ncat+1) matrix of data

  - res('yhat') = (nobs x ncat+1) matrix of fitted values probabilities: [P_0 P_1 … P_ncat] where P_j = [P_1j ; P_2j ; … ; P_nobsj]

  - res('llike') = unrestricted log likelihood

  - res('lratio') = LR test statistic against intercept-only model (all bets=0), distributed chi-squared with (nvar-1)*ncat degrees of freedom

  - res('nobs') = number of observations

  - res('nvar') = number of variables

  - res('ncat') = number of categories of dependent variable (including the reference category j = 0)

  - res('count') = vector of counts of each value taken by y, i.e., couunt = [#y=0 #y=1 … #y=ncat]

  - res('r2mf') = McFadden pseudo-R^2

  - res('rsqr') = Estrella pseudo-R^2

DESCRIPTION

Computes multivariate logit Regression.

EXAMPLE

res = multilogit1(y,x,grocer_init,grocer_maxit,grocer_tol)
 
 
Example taken from multilogit. 

               

AUTHOR

Eric Dubois 2007