<< factor_density_lko Factor Analysis fac_kalman >>
Grocer >> Factor Analysis > factor_density_lko1

factor_density_lko1

Factor density estimation using leave k-out

CALLING SEQUENCE

[res] = factor_density_lko1(Z,meth,nb_draws,tau)

PARAMETERS

Input

* Z = a (nobs x k) matrix, collecting the variables

* meth = 'stan' if the user wants to perform the analysis on standardized variables

* nb_draws = an integer, number of resamples

* tau = Rejection probability (OPTIONAL: default=0.1)

Output

* res = a results tlist, with:

   - res('meth') = 'factor density via leave k-out resampling'

   - res('y') = the (nobs x k) matrix collecting the variables

   - res('standardization') = a boolean, indicating whether variables have been standardized

   - res('# of draws') = an integrer, the # of draws

   - res('rejection prob.') = a real, the rejection probability

   - res('factors') = a (nobs x nb_draws) matrix, the factor density

   - res('factors percentiles') = a (nobs x 5) matrix, the percentiles (2.5, 25, 50, 75 and 97.5)

   - res('k out resampling') = a (nobs x nb_draws) matrix, the number of variables kept at each draw

DESCRIPTION

Factor density resampling using the leave k-out method (rejection prob.=tau).

EXAMPLE

global GROCERDIR 
load(GROCERDIR+'\data\esp_yields.dat')
X=esp_yields('series')
res = fac_density_lko(X,'stan',1000,0.05)
// provides the result of the resampling by the leave k-out methods on factors estimation on Spanisk yields data, standardized, with 1000 draws and a 0.05 rejection probability

AUTHOR

Éric Dubois 2022

Report an issue
<< factor_density_lko Factor Analysis fac_kalman >>