stepan
/
dynare
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

minor updates for identification part 

git-svn-id: file:///home/sebastien/dynare/gsa_dyn@114 f1850c17-3b45-254b-b221-fcb05880fee1
time-shift
rattoma 2009-09-30 13:36:04 +00:00
parent 6364099122
commit a8bf833095
6 changed files with 83 additions and 55 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
latex/dyn_tbx.dvi
View File

Binary file not shown.

28
latex/dyn_tbx.log
View File

@ -1,4 +1,4 @@
This is pdfeTeX, Version 3.141592-1.30.6-2.2 (MiKTeX 2.5) (preloaded format=latex 2006.9.25) 29 SEP 2009 16:45
This is pdfeTeX, Version 3.141592-1.30.6-2.2 (MiKTeX 2.5) (preloaded format=latex 2006.9.25) 30 SEP 2009 15:07
entering extended mode
**C:/SVN/dyn_util/latex/dyn_tbx.tex
(C:\SVN\dyn_util\latex\dyn_tbx.tex
@ -209,64 +209,64 @@ Overfull \hbox (26.51085pt too wide) in paragraph at lines 217--237
[]
[5]
Overfull \hbox (102.93791pt too wide) in paragraph at lines 256--271
Overfull \hbox (44.59268pt too wide) in paragraph at lines 256--271
[][][]
[]
[6] [7]
Underfull \hbox (badness 10000) in paragraph at lines 284--292
Underfull \hbox (badness 10000) in paragraph at lines 292--300
[]
Underfull \hbox (badness 10000) in paragraph at lines 284--292
Underfull \hbox (badness 10000) in paragraph at lines 292--300
[]
[8]
Overfull \hbox (12.08382pt too wide) in paragraph at lines 356--358
Overfull \hbox (12.08382pt too wide) in paragraph at lines 364--366
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_rmse_post*.fig\OT1/cmr/m/n/12 : save re-su
lts for the anal-y-sis us-ing Metropo-
[]
Overfull \hbox (8.24295pt too wide) in paragraph at lines 368--371
Overfull \hbox (8.24295pt too wide) in paragraph at lines 376--379
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_rmse_prior_<name of observedseries>_corr_*
.fig\OT1/cmr/m/n/12 : for
[]
[9]
Overfull \hbox (5.72563pt too wide) in paragraph at lines 377--382
Overfull \hbox (5.72563pt too wide) in paragraph at lines 385--390
\OT1/cmr/m/it/12 in red \OT1/cmr/m/n/12 the cdf of the log-posterior cor-re-spo
nd-ing to the best 10% RMSE's,
[]
Overfull \hbox (0.88243pt too wide) in paragraph at lines 387--390
Overfull \hbox (0.88243pt too wide) in paragraph at lines 395--398
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_rmse_prior_lik_SA_*.fig\OT1/cmr/m/n/12 : w
hen []\OT1/cmtt/m/n/12 lik_only=1\OT1/cmr/m/n/12 , this shows
[]
Overfull \hbox (7.05743pt too wide) in paragraph at lines 390--393
Overfull \hbox (7.05743pt too wide) in paragraph at lines 398--401
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_rmse_prior_post_SA_*.fig\OT1/cmr/m/n/12 :
when []\OT1/cmtt/m/n/12 lik_only=1\OT1/cmr/m/n/12 , this shows
[]
Overfull \hbox (9.38109pt too wide) in paragraph at lines 406--412
Overfull \hbox (9.38109pt too wide) in paragraph at lines 414--420
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_redform_<endo name>_vs_shocks_*.fig\OT1/cm
r/m/n/12 : shows bar charts
[]
Overfull \hbox (26.57321pt too wide) in paragraph at lines 418--423
Overfull \hbox (26.57321pt too wide) in paragraph at lines 426--431
[][]\OT1/cmr/m/n/12 Detailed re-sults of the anal-y-ses are shown in the sub-fo
lder []\OT1/cmtt/m/n/12 \GSA\redform_stab\OT1/cmr/m/n/12 ,
[]
[10]
Overfull \hbox (9.38109pt too wide) in paragraph at lines 456--461
Overfull \hbox (9.38109pt too wide) in paragraph at lines 464--469
[][][]\OT1/cmtt/m/n/12 <DYNARE_file>_redform_<endo name>_vs_shocks_*.fig\OT1/cm
r/m/n/12 : shows bar charts
[]
@ -279,6 +279,6 @@ Here is how much of TeX's memory you used:
5051 multiletter control sequences out of 60000
14651 words of font info for 57 fonts, out of 1000000 for 2000
14 hyphenation exceptions out of 8191
29i,13n,24p,337b,294s stack positions out of 5000i,500n,10000p,200000b,32768s
29i,13n,24p,335b,294s stack positions out of 5000i,500n,10000p,200000b,32768s
Output written on dyn_tbx.dvi (12 pages, 31224 bytes).
Output written on dyn_tbx.dvi (12 pages, 31748 bytes).

BIN
latex/dyn_tbx.pdf
View File

Binary file not shown.

66
latex/dyn_tbx.ps
View File

@ -1,7 +1,7 @@
%!PS-Adobe-2.0
%%Creator: dvips(k) 5.95b Copyright 2005 Radical Eye Software
%%Title: dyn_tbx.dvi
%%CreationDate: Tue Sep 29 16:45:44 2009
%%CreationDate: Wed Sep 30 15:07:48 2009
%%Pages: 12
%%PageOrder: Ascend
%%BoundingBox: 0 0 595 842
@ -12,7 +12,7 @@
%DVIPSWebPage: (www.radicaleye.com)
%DVIPSCommandLine: dvips.exe dyn_tbx.dvi
%DVIPSParameters: dpi=600
%DVIPSSource: TeX output 2009.09.29:1645
%DVIPSSource: TeX output 2009.09.30:1507
%%BeginProcSet: tex.pro 0 0
%!
/TeXDict 300 dict def TeXDict begin/N{def}def/B{bind def}N/S{exch}N/X{S
@ -3163,7 +3163,7 @@ b(Analysis)h(T)-13 b(o)t(olb)t(o)l(x)52 b(for)h(D)l(YNARE)1549
m(t)j(Researc)m(h)g(Cen)m(tre)1132 1570 y(TP361,)e(IPSC,)i(via)g(E.)g
(F)-10 b(ermi)38 b(1)1296 1720 y(21020)e(Ispra)j(\(V)-13
b(A\))37 b(Italy)1303 1869 y Fr(marco.ratto@jr)o(c.)o(it)2449
1826 y Fq(\003)1376 2103 y Fs(Septem)m(b)s(er)g(29,)h(2009)1676
1826 y Fq(\003)1376 2103 y Fs(Septem)m(b)s(er)g(30,)h(2009)1676
2578 y Fp(Abstract)501 2746 y Fo(The)32 b(Sensitivit)m(y)h(Analysis)g
(T)-8 b(o)s(olb)s(o)m(x)33 b(for)f(D)m(YNARE)h(is)g(a)f(set)h(of)f(MA)
-8 b(TLAB)501 2859 y(routines)27 b(for)g(the)g(analysis)g(of)g(DSGE)g
@ -3419,21 +3419,21 @@ g(based)g(on)257 3777 y(theoretical)36 b(momen)m(ts)g(is)g(p)s
(erformed.)50 b(Sensitivit)m(y)38 b(plots)d(are)g(pro)m(vided)h(that)f
(allo)m(w)257 3897 y(to)e(infer)g(whic)m(h)h(parameters)f(are)g(most)g
(lik)m(ely)h(to)e(b)s(e)h(less)h(iden)m(ti\014able.)404
4017 y(Prerequisite)48 b(for)d(prop)s(erly)h(running)h(all)e(the)h
(iden)m(ti\014cation)i(routines,)i(is)c(the)257 4138
y(k)m(eyw)m(ord)38 b Fj(identification;)h Fl(in)d(the)g(D)m(YNARE)f(mo)
s(del)h(\014le.)53 b(This)36 b(k)m(eyw)m(ord)i(trig-)257
4258 y(gers)44 b(the)g(computation)g(of)f(analytic)h(deriv)-5
b(ativ)m(es)46 b(of)d(the)h(mo)s(del)g(with)g(resp)s(ect)h(to)257
4378 y(estimated)39 b(parameters)e(and)g(sho)s(c)m(ks.)58
b(This)38 b(is)g(required)g(for)e(options)i Fj(morris=0,2)257
4499 y Fl(b)s(elo)m(w,)c(whic)m(h)g(implemen)m(t)h(Iskrev)f(\(2009\))e
(iden)m(ti\014cation)i(analysis.)1852 5637 y(6)p eop
4017 y(Prerequisite)26 b(for)d(prop)s(erly)h(running)h(all)e(the)h
(iden)m(ti\014cation)h(routines,)i(is)d(the)g(k)m(ey-)257
4138 y(w)m(ord)37 b Fj(identification;)j Fl(in)c(the)h(D)m(YNARE)f(mo)s
(del)h(\014le.)54 b(This)38 b(k)m(eyw)m(ord)g(triggers)257
4258 y(the)43 b(computation)g(of)f(analytic)h(deriv)-5
b(ativ)m(es)45 b(of)d(the)h(mo)s(del)g(with)g(resp)s(ect)h(to)e(esti-)
257 4378 y(mated)33 b(parameters)f(and)g(sho)s(c)m(ks.)45
b(This)33 b(is)g(required)g(for)e(options)h Fj(morris=2)i
Fl(b)s(elo)m(w,)257 4499 y(whic)m(h)g(implemen)m(t)h(Iskrev)g(\(2009\))
c(iden)m(ti\014cation)j(analysis.)1852 5637 y(6)p eop
end
%%Page: 7 7
TeXDict begin 7 6 bop 640 557 a Fl(option)33 b(name)p
1219 594 4 121 v 100 w(default)p 1614 594 V 100 w(description)p
404 597 3947 4 v 454 681 a Fj(identification)p 1219 717
404 597 3462 4 v 454 681 a Fj(identification)p 1219 717
4 121 v 103 w Fl(0)p 1614 717 V 346 w(0)f(=)h(no)f(iden)m(ti\014cation)
i(analysis)p 1219 838 V 1614 838 V 1666 802 a(1)e(=)h(p)s(erforms)g
(iden)m(ti\014cation)h(analysis:)p 1219 958 V 1614 958
@ -3442,26 +3442,34 @@ Fj(morris)p Fl(=1)864 1042 y Fj(morris)p 1219 1078 V
101 w Fl(1)p 1614 1078 V 346 w(1)f(=)h(Screening)g(analysis)h(\(T)m(yp)
s(e)g(I)s(I)f(error\))p 1219 1199 V 1614 1199 V 1666
1163 a(0)f(=)h(ANO)m(V)-11 b(A)33 b(Mapping)g(\(T)m(yp)s(e)h(I)f
(error\))f(and)h(analytic)g(deriv)-5 b(ativ)m(es)p 1219
1319 V 1614 1319 V 1666 1283 a(2)32 b(=)h(Analytic)g(deriv)-5
b(ativ)m(es)35 b(\(similar)e(to)f(T)m(yp)s(e)i(I)s(I)f(error\))607
1403 y Fj(morris_nliv)p 1219 1440 V 103 w Fl(6)p 1614
1440 V 346 w(n)m(um)m(b)s(er)h(of)e(lev)m(els)j(in)d(Morris)i(design)
607 1524 y Fj(morris_ntra)p 1219 1560 V 103 w Fl(20)p
1614 1560 V 297 w(n)m(um)m(b)s(er)g(of)e(tra)5 b(jectories)34
b(in)e(Morris)i(design)607 1644 y Fj(trans_ident)p 1219
1680 V 103 w Fl(0)p 1614 1680 V 346 w(data)e(transformation)g(for)g
(ANO)m(V)-11 b(A)34 b(Mapping)p 1219 1801 V 1614 1801
(error\))p 1219 1319 V 1614 1319 V 1666 1283 a(2)f(=)h(Analytic)g
(deriv)-5 b(ativ)m(es)35 b(\(similar)e(to)f(T)m(yp)s(e)i(I)s(I)f
(error\))607 1403 y Fj(morris_nliv)p 1219 1440 V 103
w Fl(6)p 1614 1440 V 346 w(n)m(um)m(b)s(er)h(of)e(lev)m(els)j(in)d
(Morris)i(design)607 1524 y Fj(morris_ntra)p 1219 1560
V 103 w Fl(20)p 1614 1560 V 297 w(n)m(um)m(b)s(er)g(of)e(tra)5
b(jectories)34 b(in)e(Morris)i(design)607 1644 y Fj(trans_ident)p
1219 1680 V 103 w Fl(0)p 1614 1680 V 346 w(data)e(transformation)g(for)
g(ANO)m(V)-11 b(A)34 b(Mapping)p 1219 1801 V 1614 1801
V 1666 1765 a(0)e(=)h(no)f(transformation)p 1219 1921
V 1614 1921 V 1666 1885 a(1)g(=)h(log-transformation)p
1219 2042 V 1614 2042 V 1666 2005 a(2)f(=)h(rank-transformation)p
404 2045 3947 4 v 257 2359 a(F)-8 b(or)32 b(example,)i(the)f(follo)m
404 2045 3462 4 v 257 2359 a(F)-8 b(or)32 b(example,)i(the)f(follo)m
(wing)g(commands)h(in)f(the)g(D)m(YNARE)g(mo)s(del)g(\014le)257
2715 y Fj(identification;)257 2836 y(dynare_sensitivity\(ident)q(ifi)q
(cati)q(on=1)q(,)57 b(morris=2\);)257 3192 y Fl(trigger)33
b(the)g(iden)m(ti\014cation)h(analysis)g(using)f(analytic)g(deriv)-5
b(ativ)m(es)35 b(\(Iskrev,)f(2009\).)1852 5637 y(7)p
eop end
(cati)q(on=1)q(,)57 b(morris=2\);)257 3192 y Fl(trigger)42
b(the)g(iden)m(ti\014cation)h(analysis)g(using)f(analytic)h(deriv)-5
b(ativ)m(es)43 b(\(Iskrev,)h(2009\),)257 3313 y(join)m(tly)34
b(with)f(the)g(mapping)g(of)f(the)h(acceptable)h(region.)404
3433 y(The)j(iden)m(ti\014cation)h(analysis)g(with)f(deriv)-5
b(ativ)m(es)38 b(can)f(also)f(b)s(e)h(triggered)f(b)m(y)i(the)257
3553 y(commands)257 3910 y Fj(identification;)257 4030
y(dynare_identification;)404 4387 y Fl(This)j(do)s(es)f(not)g(do)g(the)
g(mapping)g(of)g(acceptable)h(regions)f(for)g(the)g(mo)s(del)g(and)257
4507 y(uses)j(the)f(standard)g(random)g(sampler)h(of)e(D)m(YNARE.)h(It)
g(completely)h(o\013sets)f(an)m(y)257 4628 y(use)34 b(of)e(the)h
(sensitivit)m(y)i(analysis)f(to)s(olb)s(o)m(x.)1852 5637
y(7)p eop end
%%Page: 8 8
TeXDict begin 8 7 bop 257 573 a Fk(4)161 b(Directory)53
b(structure)257 792 y Fl(Sensitivit)m(y)36 b(analysis)e(results)g(are)e

16
latex/dyn_tbx.tex
View File

@ -246,10 +246,10 @@ performs all the analyses required and displays results.
Setting the option \verb"identification=1", an identification
analysis based on theoretical moments is performed. Sensitivity plots are provided that
allow to infer which parameters are most likely to be less
identifiable.
identifiable.
Prerequisite for properly running all the identification routines, is the keyword
\verb"identification;" in the DYNARE model file. This keyword triggers the computation of analytic derivatives of the model with respect to estimated parameters and shocks. This is required for options \verb"morris=0,2" below, which implement \cite{Iskrev2009} identification analysis.
\verb"identification;" in the DYNARE model file. This keyword triggers the computation of analytic derivatives of the model with respect to estimated parameters and shocks. This is required for options \verb"morris=2" below, which implement \cite{Iskrev2009} identification analysis.
\vspace{1cm}
@ -259,7 +259,7 @@ Prerequisite for properly running all the identification routines, is the keywor
& & 1 = performs identification analysis:\\
& & this forces \verb"redform"=0 and default\verb"morris"=1\\
\verb"morris"& 1 & 1 = Screening analysis (Type II error)\\
& & 0 = ANOVA Mapping (Type I error) and analytic derivatives\\
& & 0 = ANOVA Mapping (Type I error)\\
& & 2 = Analytic derivatives (similar to Type II error)\\
\verb"morris_nliv"& 6 & number of levels in Morris design\\
\verb"morris_ntra"& 20& number of trajectories in Morris design\\
@ -277,8 +277,16 @@ Prerequisite for properly running all the identification routines, is the keywor
\verb"dynare_sensitivity(identification=1, morris=2);"
\vspace{1cm}
\noindent trigger the identification analysis using analytic derivatives \citep{Iskrev2009}.
\noindent trigger the identification analysis using analytic derivatives \citep{Iskrev2009}, jointly with the mapping of the acceptable region.
The identification analysis with derivatives can also be triggered by the commands
\vspace{1cm}
\noindent\verb"identification;"\\
\verb"dynare_identification;"
\vspace{1cm}
This does not do the mapping of acceptable regions for the model and uses the standard random sampler of DYNARE. It completely offsets any use of the sensitivity analysis toolbox.
\newpage
\section{Directory structure}
Sensitivity analysis results are saved on the hard-disk of the

28
latex/dyn_tbx.tex.bak
View File

@ -29,8 +29,8 @@ TP361, IPSC, via E. Fermi 1\\21020 Ispra
\texttt{marco.ratto@jrc.it}
\thanks{The author gratefully thanks Christophe Planas, Kenneth Judd, Michel Juillard,
Alessandro Rossi, Frank Schorfheide and the participants to the
First Course on Global Sensitivity Analysis for Macroeconomic
Models (Ispra, 16-17 March 2006) for interesting discussions and
Courses on Global Sensitivity Analysis for Macroeconomic
Models (Ispra, 2006-2007-2008) for interesting discussions and
helpful suggestions.}}
%%% To have the current date inserted, use \date{\today}:
@ -52,8 +52,7 @@ the DYNARE v4 environment.
\begin{description}
\item \textbf{Keywords}: Stability Mapping , Reduced form solution, DSGE models,
Monte Carlo filtering,
Global Sensitivity Analysis, High Dimensional Model Representation.
Monte Carlo filtering, Global Sensitivity Analysis.
\end{description}
\end{abstract}
\newpage
@ -245,10 +244,14 @@ performs all the analyses required and displays results.
\subsection{Identification analysis (under development)}
Setting the option \verb"identification=1", an identification
analysis based on theoretical moments, steady state gains and time
of adjustment is performed. Sensitivity plots are provided that
analysis based on theoretical moments is performed. Sensitivity plots are provided that
allow to infer which parameters are most likely to be less
identifiable. \vspace{1cm}
identifiable.
Prerequisite for properly running all the identification routines, is the keyword
\verb"identification;" in the DYNARE model file. This keyword triggers the computation of analytic derivatives of the model with respect to estimated parameters and shocks. This is required for options \verb"morris=0,2" below, which implement \cite{Iskrev2009} identification analysis.
\vspace{1cm}
\begin{tabular}{r|l|l}
option name & default & description \\ \hline
@ -256,7 +259,7 @@ identifiable. \vspace{1cm}
& & 1 = performs identification analysis:\\
& & this forces \verb"redform"=0 and default\verb"morris"=1\\
\verb"morris"& 1 & 1 = Screening analysis (Type II error)\\
& & 0 = ANOVA Mapping (Type I error) and analytic derivatives\\
& & 0 = ANOVA Mapping (Type I error)\\
& & 2 = Analytic derivatives (similar to Type II error)\\
\verb"morris_nliv"& 6 & number of levels in Morris design\\
\verb"morris_ntra"& 20& number of trajectories in Morris design\\
@ -266,6 +269,15 @@ identifiable. \vspace{1cm}
& & 2 = rank-transformation\\\hline
\end{tabular}
\vspace{1cm}
\noindent For example, the following commands in the DYNARE model file
\vspace{1cm}
\noindent\verb"identification;"\\
\verb"dynare_sensitivity(identification=1, morris=2);"
\vspace{1cm}
\noindent trigger the identification analysis using analytic derivatives \citep{Iskrev2009}.
\newpage
\section{Directory structure}