Diff for "SbvarOptions"

Differences between revisions 24 and 25

Table Commands:

Variable	File	Line number	Option name	Comments	Defaults
q_m	msstart_setup.m	11	freq	quarters or months	4
mu	msstart_setup.m		coefficients_prior_hyperparameters (mu set automatically, option not used)	if freq is not monthly or quarterly, this option must be set
yrBin	msstart_setup.m	12	initial_year	beginning of the year
qmBin	msstart_setup.m	13	initial_subperiod	beginning of the quarter or month	1
yrFin	msstart_setup.m	14	final_year	final year
qmFin	msstart_setup.m	15	final_subperiod	final month or quarter
xdd	msstart_setup.m	22	data	Load data
vlist	msstart_setup.m	36	vlist (vlist set automatically, option not used)
varlist	msstart_setup.m	37	varlist (varlist set automatically, option not used)	name of variables
nfyr	msstart_setup.m	70	forecast	number of years for forecasting	4
indxDummy	msstart_setup.m	105	bayesian_prior	1: Bayesian prior 0: no prior	1
	in .m files in subfolders		cross_restrictions	1: cross-A0-and-A+ restrictions 0: idfile_const is all we have	0
	szemldata_a_setup.m	25	nstates (not used)	Number of states	2
	szemldata_a_setup.m	22	eq_indices (not used)	Rows corresponds to equations
	szemldata_a_setup.m	26	indxScaleStates (not used)	0 (default): no scale adjustment in the prior variance for the number of states so that the prior value is the same as the constant VAR when the parameters in all states are the same. 1: allows a scale adjustment, marking the prior variance bigger by the number of states. Wrong choice because it makes the prior looser as the number of states increases.	0
	szemldata_a_setup.m	30	alpha (not used)	Alpha value for squared time-varying structural shock lambda.	1.0
	szemldata_a_setup.m	31	beta (not used)	Beta value for squared time-varying structural shock lambda.	1.0
	szemldata_a_setup.m	32	gsig2_lmd (not used)	Case 4 (no state change across variables (i) and across lags (l); only one scale factor for all lagged variables change from state to state). Normal prior variance for structural shock lambda, one scale factor for all lagged D+ in a given equation.	50²
	szemldata_a_setup.m	34	gsig2_lmdm (not used)	Case 3 (no state change across lags (l) but allows all variables for a given lag to switch states). Normal prior variance for different variables in lagged D+.	50²
	szemldata_a_setup.m	36	q_diag (not used)	Prior mean: 0.95 for monthly data (about 20 months) and 0.85 for quarterly data (about 6.7 quarters). Diagonal q in the transtion matrix and the duration is 1 over (1-q). To allow an absorbing state, we need to have 1 for all off-diagonal elements and x for the diagonal element.	0.85
	szemldata_a_setup.m	42	flat_prior (not used)	0: diagonal of transition matrix has some duration 1: flat prior where all alpha's are 1.0 in Dirichlet.	0
	msstart2.m		restriction_fname	file for restriction on time variations
	msstart2.m		nlags	Prior: number of lags	1
	msstart2.m		contemp_reduced_form	1: contemporaneous recursive reduced form 0: restricted (non-recursive) form	0
	msstart2.m		real_pseudo_forecast	1: Pseudo forecasts 0: real time forecasts	0
	msstart2.m		dummy_obs	Dummy observations	0
nCsk	msstart2.m	24	nCsk	conditional directly on shoocks	0
nstd	msstart2.m	25	nstd	number of standard deviations to cover the range in which distributions are put to bins	6
ninv	msstart2.m	26	ninv	the number of bins for grouping, say, impulse responses	1000
Indxcol	msstart2.m	27	Indxcol	a vector of random columns in which MC draws are made	size of the number of variables
IndxParR	msstart2.m	29	IndxParR	1: parameters random; 0: no randomness in parameters	1
IndxOvR	msstart2.m	31	IndxOvR	1: distributions for other variables of interest; 0: no distribution. Example: joint distribution of a(1) and a(2). Only for specific purposes	0
Aband	msstart2.m	33	Aband	1: error bands with only A0 and A+ random.	1
IndxAp	msstart2.m	34	IndxAp	1: generate draws of A+; 0: no such draws. Note: when IndxAp=0, there is no effect from the values of IndxImf, IndxFore,or Apband.	1
Apband	msstart2.m	37	Apband	1: error bands for A+; 0: no error bands for A+.	1
IndxImf	msstart2.m	39	IndxImf	Used only if IndxAp=1. 1: generate draws of impulse responses; 0: no such draws (thus no effect from Imfband)	1
Imfband	msstart2.m	41	Imfband	Used only if IndxAp=1. 1: error bands for impulse responses; 0: no error bands	1
IndxFore	msstart2.m	42	IndxFore	Used only if IndxAp=1. 1: generate draws of forecasts; 0: no such draws (thus no effect from Foreband)	0
Foreband	msstart2.m	43	Foreband	Used only if IndxAp=1. 1: error bands for out-of-sample forecasts; 0: no error bands	0
indxGforehat	msstart2.m	45	indxGforehat	Used only if IndxAp=1. 1: plot unconditoinal forecasts; 0: no such plot	1
rnum	msstart2.m	46	rnum	Used only if IndxAp=1. Number of rows in the graph	number of variables
cnum	msstart2.m	47	cnum	Used only if IndxAp=1. Number of columns in the graph	1
indxGimfhat	msstart2.m	53	indxGimfhat	Used only if IndxAp=1. 1: plot ML impulse responses; 0: no plot	1
indxEstima	msstart2.m	54	indxEstima	Used only if IndxAp=1. 1: ML estimation; 0: no estimation and data only	1
indxgDLS	msstart2.m	71	indxgDLS	Hard conditions directly on variables. Used only if IndxAp=1. 1: graph point forecast on variables; 0: disable	1
nconstr	msstart2.m	74	nconstr	Hard conditions directly on variables. q: 4 years -- 4*12 months. When 0, no conditions directly on variables	nconstr1+nconstr2
nconstr	msstart2.m	76	nconstr	Hard conditions directly on variables. 6*nconstr1	0
eq_ms	msstart2.m	77	eq_ms	Hard conditions directly on variables. Location of MS equation; if [], all shocks
PorR	msstart2.m	78	PorR	The variable conditioned. 1: Pcm; 3: FFR; 4: CPI
Cms	msstart2.m	87	Cms	Conditions directly on future shocks. Used only if IndxAp=1. 1: condition on ms shocks; 0: disable this and "fidcnderr.m" gives unconditional forecasts if nconstr = 0 as well	0
nCms	msstart2.m	89	nCms	Conditions directly on future shocks. Used only if IndxAp=1. Number of the stance of policy; 0 if no tightening or loosening	0
eq_Cms	msstart2.m	90	eq_Cms	Conditions directly on future shocks. Used only if IndxAp=1. Location of MS shocks	1
TLindx	msstart2.m	91	TLindx	Conditions directly on future shocks. Used only if IndxAp=1. 1-by-nCms vector; 1: tightening; 0: loosen
TLnumber	msstart2.m	92	TLnumber	Conditions directly on future shocks. Used only if IndxAp=1. 1-by-nCms vector; cut-off point for MS shocks
nbancon	msstart2.m	120	cnum	Soft conditions on variables. Number of band conditions; when 0, disable this option Note that each condition corresponds to variable	0
banact	msstart2.m	122	banact	Soft conditions on variables. 1: use infor on actual; 0: preset without infor on actual	1
output_file_tag	sz_prd.m	24	output_file_tag	tag for output files	test_2v
create_initialization_file	sz_prd.m	27	create_initialization_file	1 to create init_<output_file_tag>.dat, 0 otherwise	1
estimate_msmodel	sz_prd.m	30	estimate_msmodel	1 to perform estimation, 0 otherwise	1
compute_mdd	sz_prd.m	33	compute_mdd	1 to perform estimation, 0 otherwise	1
compute_probabilities	sz_prd.m	36	compute_probabilities	1 to compute probabilites, 0 otherwise	1
print_draws	sz_prd.m	39	print_draws	1 to Prints draws of the posterior, 0 otherwise	1
n_draws	sz_prd.m	40	n_draws	Number of draws from the posterior	1000
thinning_factor	sz_prd.m	41	thinning_factor	Thinning Factor	1
markov_file	sz_prd.m	46	markov_file	File should be specified by the user. If not specified, it will use the Sims, Waggoner, and Tao's specification	specification_2v2c.dat
mhm_file	sz_prd.m	52	mhm_file	File should be specified by the user. If not specified, it will use the Sims, Waggoner, and Tao's specification	MHM_input.dat
proposal_draws	sz_prd.m	53	proposal_draws	Proposal draws for the Markov specification file	100000
firstMetrop	generateMHM_input.m	8	draws_nbr_burn_in_1	?	30000
scondMetrop	generateMHM_input.m	15	draws_nbr_burn_in_2	?	10000
ndrawsmv	generateMHM_input.m	21	draws_nbr_mean_var_estimate	?	200000
ndrawsmhm	generateMHM_input.m	27	draws_nbr_modified_harmonic_mean	?	1000000
tfmhm	generateMHM_input.m	33	thinning_factor	?	1
svd	generateMHM_input.m	40	dirichlet_scale	?	3
opt1	generateMHM_input.m	44	?	to be hardcoded (maybe not used)	1.0
opt2	generateMHM_input.m	46	?	to be hardcoded (maybe not used)	1.5
opt3	generateMHM_input.m	48	?	to be hardcoded (maybe not used)	2.0

-  ⇤ ← Revision 24 as of 2011-05-30 15:14:17 → 
  Size: 11014
  Editor: HoutanBastani
  Comment:
+   ← Revision 25 as of 2011-05-31 12:23:59 → ⇥
  Size: 11104
  Editor: HoutanBastani
  Comment:
-Deletions are marked like this.
+Additions are marked like this.
 Line 7:
-|| mu  |||| msstart_setup.m ||||  |||| coefficients_prior_hyperparameters |||| ~- if freq is not monthly or quarterly, this option must be set -~ ||||  ||
+|| mu  |||| msstart_setup.m ||||  |||| coefficients_prior_hyperparameters (mu set automatically, option not used) |||| ~- if freq is not monthly or quarterly, this option must be set -~ ||||  ||
 Line 13:
-|| vlist |||| msstart_setup.m |||| 36 ||||vlist |||| |||| ||
|| varlist |||| msstart_setup.m |||| 37 ||||varlist ||||~- name of variables -~|||| ||
|| idfile_const |||| msstart_setup.m |||| 40 ||||restriction_fname |||| ~-file for restriction on time variations-~ |||| ||
|| nfyr |||| msstart_setup.m |||| 70 ||||forecast ||||~- number of years for forecasting-~ |||| 1 ||
|| lags |||| msstart_setup.m |||| 97 ||||nlags |||| ~- Prior: number of lags-~ |||| ||
|| indxc0Pres |||| msstart_setup.m |||| 98 ||||cross_restrictions ||||~- '''1''': cross-A0-and-A+ restrictions <<BR>> '''0''': idfile_const is all we have-~ |||| 0 ||
|| Rform |||| msstart_setup.m |||| 100 ||||contemp_reduced_form |||| ~-'''1''': contemporaneous recursive reduced form <<BR>> '''0''': restricted (non-recursive) form -~|||| 0 ||
|| Pseudo |||| msstart_setup.m |||| 101 ||||real_pseudo_forecast |||| ~-'''1''': Pseudo forecasts <<BR>> '''0''': real time forecasts-~ |||| 0 ||
|| indxPrior |||| msstart_setup.m |||| 102 ||||bayesian_prior |||| ~-'''1''': Bayesian prior <<BR>> '''0''': no prior-~ |||| 1 ||
|| ndobs |||| msstart_setup.m |||| 104 ||||dummy_obs |||| ~-Dummy observations -~|||| 0 ||
||nStates||||szemldata_a_setup.m||||25||||nstates||||~-Number of states-~ ||||2      ||
||indxEqnTv_m||||szemldata_a_setup.m||||22||||eq_indices||||~-Rows corresponds to equations-~ ||||      ||
||indxScaleStates||||szemldata_a_setup.m||||26||||indxScaleStates||||~- '''0''' (default): no scale adjustment in the prior variance for the number of states so that the prior value is the same as the constant VAR when the parameters in all states are the same.<<BR>> '''1''': allows a scale adjustment, marking the prior variance bigger by the number of states. Wrong choice because it makes the prior looser as the number of states increases.-~  ||||0      ||
||galp||||szemldata_a_setup.m||||30||||alpha||||~-Alpha value for squared time-varying structural shock lambda.-~ ||||1.0      ||
||gbeta||||szemldata_a_setup.m||||31||||beta||||~-Beta value for squared time-varying structural shock lambda.-~ ||||1.0      ||
||gsig2_lmd||||szemldata_a_setup.m||||32||||gsig2_lmd||||~-Case 4 (no state change across variables (i) and across lags (l); only one scale factor for all lagged variables change from state to state).   Normal prior variance for structural shock lambda, one scale factor for all lagged D+ in a given equation.-~ ||||50^2^      ||
||gsig2_lmdm||||szemldata_a_setup.m||||34||||gsig2_lmdm||||~-Case 3 (no state change across lags (l) but allows all variables for a given lag to switch states). Normal prior variance for different variables in lagged D+.-~ ||||50^2^      ||
||q_diag||||szemldata_a_setup.m||||36||||q_diag||||~-Prior mean: 0.95 for monthly data (about 20 months) and 0.85 for quarterly data (about 6.7 quarters). Diagonal q in the transtion matrix and the duration is 1 over (1-q). To allow an absorbing state, we need to have 1 for all off-diagonal elements and x for the diagonal element.-~ ||||0.85      ||
||indxFlatPriorP||||szemldata_a_setup.m||||42||||flat_prior||||~-'''0''': diagonal of transition matrix has some duration <<BR>> '''1''': flat prior where all alpha's are 1.0 in Dirichlet.-~ ||||0      ||
+|| vlist |||| msstart_setup.m |||| 36 ||||vlist (vlist set automatically, option not used) |||| |||| ||
|| varlist |||| msstart_setup.m |||| 37 ||||varlist (varlist set automatically, option not used) ||||~- name of variables -~|||| ||
|| nfyr |||| msstart_setup.m |||| 70 ||||forecast ||||~- number of years for forecasting-~ |||| 4 ||
|| indxDummy |||| msstart_setup.m |||| 105 ||||bayesian_prior |||| ~-'''1''': Bayesian prior <<BR>> '''0''': no prior-~ |||| 1 ||
||  |||| in .m files in subfolders ||||  ||||cross_restrictions ||||~- '''1''': cross-A0-and-A+ restrictions <<BR>> '''0''': idfile_const is all we have-~ |||| 0 ||
||  ||||szemldata_a_setup.m||||25||||nstates (not used) ||||~-Number of states-~ ||||2      ||
|| ||||szemldata_a_setup.m||||22||||eq_indices (not used) ||||~-Rows corresponds to equations-~ ||||      ||
|| ||||szemldata_a_setup.m||||26||||indxScaleStates (not used) ||||~- '''0''' (default): no scale adjustment in the prior variance for the number of states so that the prior value is the same as the constant VAR when the parameters in all states are the same.<<BR>> '''1''': allows a scale adjustment, marking the prior variance bigger by the number of states. Wrong choice because it makes the prior looser as the number of states increases.-~  ||||0      ||
|| ||||szemldata_a_setup.m||||30||||alpha (not used)||||~-Alpha value for squared time-varying structural shock lambda.-~ ||||1.0      ||
|| ||||szemldata_a_setup.m||||31||||beta (not used)||||~-Beta value for squared time-varying structural shock lambda.-~ ||||1.0      ||
|| ||||szemldata_a_setup.m||||32||||gsig2_lmd (not used)||||~-Case 4 (no state change across variables (i) and across lags (l); only one scale factor for all lagged variables change from state to state).   Normal prior variance for structural shock lambda, one scale factor for all lagged D+ in a given equation.-~ ||||50^2^      ||
|| ||||szemldata_a_setup.m||||34||||gsig2_lmdm (not used)||||~-Case 3 (no state change across lags (l) but allows all variables for a given lag to switch states). Normal prior variance for different variables in lagged D+.-~ ||||50^2^      ||
|| ||||szemldata_a_setup.m||||36||||q_diag (not used)||||~-Prior mean: 0.95 for monthly data (about 20 months) and 0.85 for quarterly data (about 6.7 quarters). Diagonal q in the transtion matrix and the duration is 1 over (1-q). To allow an absorbing state, we need to have 1 for all off-diagonal elements and x for the diagonal element.-~ ||||0.85      ||
|| ||||szemldata_a_setup.m||||42||||flat_prior (not used)||||~-'''0''': diagonal of transition matrix has some duration <<BR>> '''1''': flat prior where all alpha's are 1.0 in Dirichlet.-~ ||||0      ||
||  |||| msstart2.m ||||  ||||restriction_fname  |||| ~-file for restriction on time variations-~ |||| ||
||  |||| msstart2.m ||||  ||||nlags |||| ~- Prior: number of lags-~ |||| 1 ||
||  |||| msstart2.m ||||  ||||contemp_reduced_form |||| ~-'''1''': contemporaneous recursive reduced form <<BR>> '''0''': restricted (non-recursive) form -~|||| 0 ||
||  |||| msstart2.m ||||  ||||real_pseudo_forecast |||| ~-'''1''': Pseudo forecasts <<BR>> '''0''': real time forecasts-~ |||| 0 ||
||  |||| msstart2.m ||||  ||||dummy_obs |||| ~-Dummy observations -~|||| 0 ||