Page 1 of 1

simul_replic=100 query

PostPosted: Mon Nov 04, 2013 4:54 pm
by mauriceroche
I have a query regarding the output from the following line.

stoch_simul(order=1,simul_replic=100,periods=300,hp_filter=1600,irf=40) y c ;

MOMENTS OF SIMULATED VARIABLES (HP filter, lambda = 1600)

VARIABLE MEAN STD. DEV. VARIANCE SKEWNESS KURTOSIS
y 1.159543 0.417516 0.174319 0.357511 -0.049269
c 0.379441 0.596591 0.355921 0.363340 -0.030360

Are the moments the averages of the 100 replications of a sample of size 300-100?

Re: simul_replic=100 query

PostPosted: Tue Nov 05, 2013 9:56 am
by jpfeifer
See the manual
simul_replic = INTEGER
Number of series to simulate when empirical moments are requested (i.e. periods
> 0). Note that if this option is greater than 1, the additional series will not be used
for computing the empirical moments but will simply be saved in binary form to
the file ‘FILENAME_simul’. Default: 1.

Re: simul_replic=100 query

PostPosted: Tue Nov 05, 2013 10:54 pm
by mauriceroche
Thanks.

In the binary file FILENAME_simul, what way are variables organized?

For example in stoch_simul(order=1,simul_replic=100,periods=300,hp_filter=1600,irf=40) y c ;

Is there 300*100 simulated series for y and then 300*100 simulated series for c?
Or is it 300 observations on y, then 300 observations on c, repeated 100 times.
Are the first 100 observations removed prior to writing to FILENAME_simul?