Added tutorial examples that run in Octave

v4.1/BVAR tutorial Octave/example_0_minn.m

@@ -0,0 +1,54 @@
+%% BVAR tutorial: Inference with Minnesota Prior
+% Author:   Filippo Ferroni
+% Date:     27/02/2020
+
+clear all
+close all
+clc
+addpath ../../cmintools/
+addpath ../../v4.1/
+pkg load statistics
+
+%% %=========================================================================
+%%% INFERENCE %%%
+%%=========================================================================
+
+% load the data
+load('../BVAR tutorial/Data.mat')
+y= [IPI HICP CORE Euribor1Y M3 EXRATE];% collect the variables used in the VAR
+
+%% Ex1/ Minnesota Prior
+
+lags = 6; 
+options.max_minn_hyper  = 1;
+options.minn_prior_tau  = 10;      % set tau 
+options.index_est       = [3 4];   % define the hyper-parameters over which to maximize
+options.lb              = [0 0];   % sets the lower bounds 
+options.ub              = [20 20]; % sets the upper bounds
+options.max_compute     = 3;       % optimization  by Matlab Simplex
+BVAR                    = bvar(y,lags,options);
+
+%% Ex2/ Minnesota Prior
+
+clear options
+lags = 6; 
+% setting the default values for the hyperparameters
+hyperpara(1)    = 3;		  % tau
+hyperpara(2)    = 0.5;		  % decay
+hyperpara(3)    = 5;		  % lambda
+hyperpara(4)    = 2;		  % mu
+hyperpara(5)    = 2;		  % omega
+% setting the options
+options.index_est	   = 1:1;      % hyper-parameter over which maximize
+options.max_compute    = 3;      % maximize  using Matlab fmincon function
+options.lb             = [0.05]; % Lower bound
+[postmode,logmlike,HH] = bvar_max_hyper(hyperpara,y,lags,options); 
+
+
+%% Ex3
+
+hyperpara(1)            = postmode(1); % use as starting value previous mode
+options.index_est       = 1:3; % set hyper-parameters over which maximize
+options.lb              = [0.05 0.05 0.05]; % Lower bounds
+options.ub              = [50 50 50];       % Upper bounds
+[postmode,log_dnsty,HH] = bvar_max_hyper(hyperpara,y,lags,options);