Skip to content
Snippets Groups Projects
Commit 158aa2de authored by Marco Ratto's avatar Marco Ratto
Browse files

Update of gsa documentation 

(cherry picked from commit 5539910c)
parent 14173fa4
Branches
Tags
No related merge requests found
Hide whitespace changes
Inline Side-by-side
doc/gsa/gsa.tex
+ 28
− 34
View file @ 158aa2de
...@@ -32,13 +32,13 @@ ...@@ -32,13 +32,13 @@
\author{Marco Ratto\\ \author{Marco Ratto\\
European Commission, Joint Research Centre \\ European Commission, Joint Research Centre \\
TP361, IPSC, via E. Fermi 1\\21020 Ispra TP361, IPSC, \\21027 Ispra
(VA) Italy\\ (VA) Italy\\
\texttt{marco.ratto@jrc.it} \texttt{marco.ratto@jrc.ec.europa.eu}
\thanks{The author gratefully thanks Christophe Planas, Kenneth Judd, Michel Juillard, \thanks{The author gratefully thanks Christophe Planas, Kenneth Judd, Michel Juillard,
Alessandro Rossi, Frank Schorfheide and the participants to the Alessandro Rossi, Frank Schorfheide and the participants to the
Courses on Global Sensitivity Analysis for Macroeconomic Courses on Global Sensitivity Analysis for Macroeconomic
Models (Ispra, 2006-2007-2008) for interesting discussions and Models (Ispra, 2006-2007-2008-2010) for interesting discussions and
helpful suggestions.}} helpful suggestions.}}
%%% To have the current date inserted, use \date{\today}: %%% To have the current date inserted, use \date{\today}:
...@@ -123,7 +123,10 @@ beforehand}. ...@@ -123,7 +123,10 @@ beforehand}.
\verb"morris_ntra"& 20& number of trajectories in Morris design\\ \verb"morris_ntra"& 20& number of trajectories in Morris design\\
\verb"ppost"& 0& 0 = don't use Metropolis posterior sample\\ \verb"ppost"& 0& 0 = don't use Metropolis posterior sample\\
& & 1 = use Metropolis posterior sample: this \\ & & 1 = use Metropolis posterior sample: this \\
& & \hspace{0.5 cm} overrides any other sampling option! \\ \hline & & \hspace{0.5 cm} overrides any other sampling option! \\
\verb"neighborhood_width"& []& $\delta$ (real number$>0$) uniform sample in the\\
& & neighborhood of the posterior mode $\hat{\theta}$ \\
& & interval width: $\hat{\theta}(1\pm\delta)$ \\\hline
\end{tabular} \end{tabular}
\subsection{Stability mapping} \subsection{Stability mapping}
\begin{tabular}{r|l|l} \begin{tabular}{r|l|l}
...@@ -132,11 +135,13 @@ beforehand}. ...@@ -132,11 +135,13 @@ beforehand}.
& & 0 = no stability mapping is performed\\ & & 0 = no stability mapping is performed\\
\verb"load_stab"& 0& 0 = generate a new sample\\ \verb"load_stab"& 0& 0 = generate a new sample\\
& & 1 = load a previously created sample \\ & & 1 = load a previously created sample \\
\verb"alpha2_stab"& 0.4& critical value for correlations $\rho$ in filtered samples:\\ \verb"pvalue_corr"& 0.001& critical p-value for correlations $\rho$ in filtered samples:\\
& & plot couples of parameters with \\ & & plot couples of parameters with \\
& & $|\rho|>$\verb"alpha2_stab"\\ & & p-value$<$\verb"pvalue_corr"\\
\verb"ksstat"& 0.1& critical value for Smirnov statistics $d$: \\ \verb"pvalue_ks" & 0.001& critical p-value for Smirnov statistics $d$: \\
& & plot parameters with $d>$\verb"ksstat"\\ \hline & & plot parameters with p-value$<$\verb"pvalue_ks"\\
\verb"lik_init" & 1& 1 = the model is stationary (unit roots are `explosive')\\
& & 3 = the model has unit roots (unit roots are `stable')\\ \hline
\end{tabular} \end{tabular}
\newpage \newpage
...@@ -167,10 +172,13 @@ out-of-sample validation. \vspace{0.5cm} ...@@ -167,10 +172,13 @@ out-of-sample validation. \vspace{0.5cm}
& & of reduced form coefficients\\ & & of reduced form coefficients\\
& & [\verb"max" \verb"max"] = analyse filtered \\ & & [\verb"max" \verb"max"] = analyse filtered \\
& & entries within the range [\verb"max" \verb"max"] \\ & & entries within the range [\verb"max" \verb"max"] \\
\verb"ksstat_redform"& 0.001& p-value for Smirnov statistics $d$ \\ \verb"ksstat_redform"& 0.001& critical p-value for Smirnov statistics $d$ \\
& & when reduced form entries are filtered\\ & & when \verb"threshold_redform" is active\\
\verb"alpha2_redform"& 0.001& p-value for correlation $\rho$ \\ & & plot parameters with p-value$<$\verb"ksstat_redform"\\
& & when reduced form entries are filtered\\ \verb"alpha2_redform"& 0.001& critical p-value for correlation $\rho$ \\
& & when \verb"threshold_redform" is active\\
& & plot couples of parameters with \\
& & p-value$<$\verb"alpha2_redform"\\
\verb"namendo"& () & list of endogenous variables \\ \verb"namendo"& () & list of endogenous variables \\
& : & jolly character to indicate ALL endogenous \\ & : & jolly character to indicate ALL endogenous \\
\verb"namlagendo"& () & list of lagged endogenous variables:\\ \verb"namlagendo"& () & list of lagged endogenous variables:\\
...@@ -212,6 +220,7 @@ options of \verb"dynare_estimation". These are: ...@@ -212,6 +220,7 @@ options of \verb"dynare_estimation". These are:
\item \verb"datafile" \item \verb"datafile"
\item \verb"mode_file" \item \verb"mode_file"
\item \verb"first_obs" \item \verb"first_obs"
\item \verb"lik_init"
\item \verb"nobs" \item \verb"nobs"
\item \verb"prefilter" \item \verb"prefilter"
\item \verb"presample" \item \verb"presample"
...@@ -235,10 +244,10 @@ options of \verb"dynare_estimation". These are: ...@@ -235,10 +244,10 @@ options of \verb"dynare_estimation". These are:
\verb"istart_rmse"& 1& start computing RMSE's from \verb"istart_rmse":\\ \verb"istart_rmse"& 1& start computing RMSE's from \verb"istart_rmse":\\
& & use 2 to avoid big initial error \\ & & use 2 to avoid big initial error \\
\verb"alpha_rmse"& 0.001& p-value for Smirnov statistics $d$:\\ \verb"alpha_rmse"& 0.001& p-value for Smirnov statistics $d$:\\
& & plot parameters with $pvalue<$\verb"alpha_rmse"\\ & & plot parameters with p-value$<$\verb"alpha_rmse"\\
\verb"alpha2_rmse"& 0.001& p-value for correlation $\rho$\\ \verb"alpha2_rmse"& 0.001& p-value for correlation $\rho$\\
& & plot couples of parameters with & & plot couples of parameters with
$pvalue<$\verb"alpha2_rmse"\\ p-value$<$\verb"alpha2_rmse"\\
\end{tabular} \end{tabular}
\subsection{Screening analysis} \subsection{Screening analysis}
...@@ -248,14 +257,12 @@ with respect to those listed in the `Sampling options': ...@@ -248,14 +257,12 @@ with respect to those listed in the `Sampling options':
performs all the analyses required and displays results. performs all the analyses required and displays results.
\subsection{Identification analysis (under development)} \subsection{Identification analysis}
Setting the option \verb"identification=1", an identification Setting the option \verb"identification=1", an identification
analysis based on theoretical moments 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 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=2" below, which implement \cite{Iskrev2009} identification analysis.
\vspace{1cm} \vspace{1cm}
...@@ -263,36 +270,23 @@ Prerequisite for properly running all the identification routines, is the keywor ...@@ -263,36 +270,23 @@ Prerequisite for properly running all the identification routines, is the keywor
option name & default & description \\ \hline option name & default & description \\ \hline
\verb"identification"& 0 & 0 = no identification analysis \\ \verb"identification"& 0 & 0 = no identification analysis \\
& & 1 = performs identification analysis:\\ & & 1 = performs identification analysis:\\
& & this forces \verb"redform"=0 and default\verb"morris"=1\\ & & this forces \verb"redform"=0 and default \verb"morris"=1\\
\verb"morris"& 1 & 1 = Screening analysis (Type II error)\\ \verb"morris"& 1 & 1 = Screening analysis (Type II error)\\
& & 0 = ANOVA Mapping (Type I error)\\ & & 2 = Analytic derivatives \citep{Iskrev2010,Iskrev2011}\\
& & 2 = Analytic derivatives (similar to Type II error)\\
\verb"morris_nliv"& 6 & number of levels in Morris design\\ \verb"morris_nliv"& 6 & number of levels in Morris design\\
\verb"morris_ntra"& 20& number of trajectories in Morris design\\ \verb"morris_ntra"& 20& number of trajectories in Morris design\\
\verb"trans_ident"& 0 & data transformation for ANOVA Mapping\\
& & 0 = no transformation\\
& & 1 = log-transformation\\
& & 2 = rank-transformation\\\hline
\end{tabular} \end{tabular}
\vspace{1cm} \vspace{1cm}
\noindent For example, the following commands in the DYNARE model file \noindent For example, the following commands in the DYNARE model file
\vspace{1cm} \vspace{1cm}
\noindent\verb"identification;"\\ \noindent\verb"dynare_sensitivity(identification=1, morris=2);"
\verb"dynare_sensitivity(identification=1, morris=2);"
\vspace{1cm} \vspace{1cm}
\noindent trigger the identification analysis using analytic derivatives \citep{Iskrev2009}, jointly with the mapping of the acceptable region. \noindent trigger the identification analysis using \cite{Iskrev2010,Iskrev2011}, 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 \newpage
\section{Directory structure} \section{Directory structure}
Sensitivity analysis results are saved on the hard-disk of the Sensitivity analysis results are saved on the hard-disk of the
......
doc/gsa/marco.bib
+ 17
− 7
View file @ 158aa2de
@TECHREPORT{Iskrev2009,
AUTHOR={Iskrev, Nikolay},
TITLE={Local Identification in \textsc{DSGE} Models},
YEAR=2009,
TYPE={mimeo},
}
@ARTICLE{Ratto_CompEcon_2008, @ARTICLE{Ratto_CompEcon_2008,
author = {Ratto, M.}, author = {Ratto, M.},
title = {Analysing DSGE Models with Global Sensitivity Analysis}, title = {Analysing DSGE Models with Global Sensitivity Analysis},
...@@ -13,3 +6,20 @@ TYPE={mimeo}, ...@@ -13,3 +6,20 @@ TYPE={mimeo},
volume = {31}, volume = {31},
pages = {115--139}, pages = {115--139},
} }
@ARTICLE{Iskrev2010,
author = {Nikolay Iskrev},
title = {Local Identification in {DSGE} Models},
journal = {Journal of Monetary Economics},
year = {2010},
volume = {57},
pages = {189-202}
}
@UNPUBLISHED{Iskrev2011,
author = {Nikolay Iskrev},
title = {Evaluating the strenght of identification in {DSGE} models. An a priori
approach},
note = {mimeo},
year = {2011}
}
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment