1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
|
# Copyright (C) 2003,2004,2005 Michael Creel <michael.creel@uab.es>
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; If not, see <http://www.gnu.org/licenses/>.
# usage: [theta, V, obj_value] =
# gmm_results(theta, data, weight, moments, momentargs, names, title, unscale, control, nslaves)
#
# inputs:
# theta: column vector initial parameters
# data: data matrix
# weight: the GMM weight matrix
# moments: name of function computes the moments
# (should return nXg matrix of contributions)
# momentargs: (cell) additional inputs needed to compute moments.
# May be empty ("")
# names: vector of parameter names
# e.g., names = char("param1", "param2");
# title: string, describes model estimated
# unscale: (optional) cell that holds means and std. dev. of data
# (see scale_data)
# control: (optional) BFGS or SA controls (see bfgsmin and samin). May be empty ("").
# nslaves: (optional) number of slaves if executed in parallel
# (requires MPITB)
#
# outputs:
# theta: GMM estimated parameters
# V: estimate of covariance of parameters. Assumes the weight matrix
# is optimal (inverse of covariance of moments)
# obj_value: the value of the GMM objective function
#
# please type "gmm_example" while in octave to see an example
function [theta, V, obj_value] = gmm_results(theta, data, weight, moments, momentargs, names, title, unscale, control, nslaves)
if nargin < 10 nslaves = 0; endif # serial by default
if nargin < 9
[theta, obj_value, convergence] = gmm_estimate(theta, data, weight, moments, momentargs, "", nslaves);
else
[theta, obj_value, convergence] = gmm_estimate(theta, data, weight, moments, momentargs, control, nslaves);
endif
m = feval(moments, theta, data, momentargs); # find out how many obsns. we have
n = rows(m);
if convergence == 1
convergence="Normal convergence";
else
convergence="No convergence";
endif
V = gmm_variance(theta, data, weight, moments, momentargs);
# unscale results if argument has been passed
# this puts coefficients into scale corresponding to the original data
if nargin > 7
if iscell(unscale)
[theta, V] = unscale_parameters(theta, V, unscale);
endif
endif
[theta, V] = delta_method("parameterize", theta, {data, moments, momentargs}, V);
k = rows(theta);
se = sqrt(diag(V));
printf("\n\n******************************************************\n");
disp(title);
printf("\nGMM Estimation Results\n");
printf("BFGS convergence: %s\n", convergence);
printf("\nObjective function value: %f\n", obj_value);
printf("Observations: %d\n", n);
junk = "X^2 test";
df = n - k;
if df > 0
clabels = char("Value","df","p-value");
a = [n*obj_value, df, 1 - chisquare_cdf(n*obj_value, df)];
printf("\n");
prettyprint(a, junk, clabels);
else
disp("\nExactly identified, no spec. test");
end;
# results for parameters
a =[theta, se, theta./se, 2 - 2*normal_cdf(abs(theta ./ se))];
clabels = char("estimate", "st. err", "t-stat", "p-value");
printf("\n");
prettyprint(a, names, clabels);
printf("******************************************************\n");
endfunction
|
