491 lines
18 KiB
Matlab
491 lines
18 KiB
Matlab
function varargout = dyn_ols(ds, fitted_names_dict, eqtags)
|
|
% function varargout = dyn_ols(ds, fitted_names_dict, eqtags)
|
|
% Run OLS on chosen model equations; unlike olseqs, allow for time t
|
|
% endogenous variables on LHS
|
|
%
|
|
% INPUTS
|
|
% ds [dseries] data
|
|
% fitted_names_dict [cell] Nx2 or Nx3 cell array to be used in naming fitted
|
|
% values; first column is the equation tag,
|
|
% second column is the name of the
|
|
% associated fitted value, third column
|
|
% (if it exists) is the function name of
|
|
% the transformation to perform on the
|
|
% fitted value.
|
|
% eqtags [cellstr] names of equation tags to estimate. If empty,
|
|
% estimate all equations
|
|
%
|
|
% OUTPUTS
|
|
% varargout{1} [dseries] data updated with fitted values (if not
|
|
% called from olsgibbs)
|
|
%
|
|
% SPECIAL REQUIREMENTS
|
|
% none
|
|
|
|
% Copyright (C) 2017-2018 Dynare Team
|
|
%
|
|
% This file is part of Dynare.
|
|
%
|
|
% Dynare 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 3 of the License, or
|
|
% (at your option) any later version.
|
|
%
|
|
% Dynare 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 Dynare. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
global M_ oo_ options_
|
|
|
|
assert(nargin >= 1 && nargin <= 3, 'dyn_ols: takes between 1 and 3 arguments');
|
|
assert(isdseries(ds), 'dyn_ols: the first argument must be a dseries');
|
|
|
|
jsonfile = [M_.fname filesep() 'model' filesep() 'json' filesep() 'modfile-original.json'];
|
|
if exist(jsonfile, 'file') ~= 2
|
|
error('Could not find %s! Please use the json=compute option (See the Dynare invocation section in the reference manual).', jsonfile);
|
|
end
|
|
|
|
%% Get Equation(s)
|
|
jsonmodel = loadjson(jsonfile);
|
|
ast = jsonmodel.abstract_syntax_tree;
|
|
jsonmodel = jsonmodel.model;
|
|
|
|
if nargin == 1
|
|
fitted_names_dict = {};
|
|
elseif nargin == 2
|
|
assert(isempty(fitted_names_dict) || ...
|
|
(iscell(fitted_names_dict) && ...
|
|
(size(fitted_names_dict, 2) == 2 || size(fitted_names_dict, 2) == 3)), ...
|
|
'dyn_ols: the second argument must be an Nx2 or Nx3 cell array');
|
|
elseif nargin == 3
|
|
ast = getEquationsByTags(ast, 'name', eqtags);
|
|
jsonmodel = getEquationsByTags(jsonmodel, 'name', eqtags);
|
|
end
|
|
|
|
%% Estimation
|
|
|
|
for i = 1:length(ast)
|
|
if ~strcmp(ast{i}.AST.node_type, 'BinaryOpNode') ...
|
|
|| ~strcmp(ast{i}.AST.op, '=')
|
|
ols_error('expecting equation with equal sign', line);
|
|
end
|
|
|
|
% Check LHS
|
|
if ~strcmp(ast{i}.AST.arg1.node_type, 'VariableNode') ...
|
|
&& ~strcmp(ast{i}.AST.arg1.node_type, 'UnaryOpNode')
|
|
ols_error('expecting Variable or UnaryOp on LHS', line);
|
|
else
|
|
if strcmp(ast{i}.AST.arg1.node_type, 'VariableNode') ...
|
|
&& (ast{i}.AST.arg1.lag ~= 0 ...
|
|
|| ~any(strcmp(ds.name, ast{i}.AST.arg1.name)))
|
|
ols_error('the lhs of the equation must be an Variable or UnaryOp with lag == 0 that exists in the dataset', line);
|
|
end
|
|
if strcmp(ast{i}.AST.arg1.node_type, 'UnaryOpNode') ...
|
|
&& (ast{i}.AST.arg1.arg.lag ~= 0 ...
|
|
|| ~any(strcmp(ds.name, ast{i}.AST.arg1.arg.name)))
|
|
ols_error('the lhs of the equation must be an Variable or UnaryOp with lag == 0 that exists in the dataset', line);
|
|
end
|
|
end
|
|
|
|
% Set LHS (Y)
|
|
lhssub = dseries();
|
|
Y = evalNode(ds, ast{i}.AST.arg1, jsonmodel{i}.line, dseries());
|
|
|
|
% Set RHS (X)
|
|
plus_node = ast{i}.AST.arg2;
|
|
last_node_to_parse = [];
|
|
residual = '';
|
|
X = dseries();
|
|
while ~isempty(plus_node) || ~isempty(last_node_to_parse)
|
|
Xtmp = dseries();
|
|
if isempty(last_node_to_parse)
|
|
[plus_node, node_to_parse, last_node_to_parse] = findNextplus_node(plus_node, jsonmodel{i}.line);
|
|
else
|
|
node_to_parse = last_node_to_parse;
|
|
last_node_to_parse = [];
|
|
end
|
|
if strcmp(node_to_parse.node_type, 'VariableNode') || strcmp(node_to_parse.node_type, 'UnaryOpNode')
|
|
if strcmp(node_to_parse.node_type, 'VariableNode')
|
|
if strcmp(node_to_parse.type, 'parameter')
|
|
% Intercept
|
|
Xtmp = dseries(ones(ds.nobs, 1), ds.firstdate, node_to_parse.name);
|
|
elseif strcmp(node_to_parse.type, 'exogenous') && ~any(strcmp(ds.name, node_to_parse.name))
|
|
% Residual if not contained in ds
|
|
if isempty(residual)
|
|
residual = node_to_parse.name;
|
|
else
|
|
ols_error(['only one residual allowed per equation; encountered ' residual ' & ' node_to_parse.name], jsonmodel{i}.line);
|
|
end
|
|
elseif strcmp(node_to_parse.type, 'endogenous') ...
|
|
|| (strcmp(node_to_parse.type, 'exogenous') && any(strcmp(ds.name, node_to_parse.name)))
|
|
% Subtract VariableNode from LHS
|
|
% NB: treat exogenous that exist in ds as endogenous
|
|
lhssub = lhssub + evalNode(ds, node_to_parse, jsonmodel{i}.line, dseries());
|
|
else
|
|
ols_error('unexpected variable type found', jsonmodel{i}.line);
|
|
end
|
|
else
|
|
% Subtract UnaryOpNode from LHS
|
|
% NB: treat exogenous that exist in ds as endogenous
|
|
lhssub = lhssub + evalNode(ds, node_to_parse, jsonmodel{i}.line, dseries());
|
|
end
|
|
elseif strcmp(node_to_parse.node_type, 'BinaryOpNode') && strcmp(node_to_parse.op, '*')
|
|
% Parse param_expr * endog_expr
|
|
Xtmp = parseTimesNode(ds, node_to_parse, jsonmodel{i}.line);
|
|
if length(Xtmp.name) > 1
|
|
% Handle constraits
|
|
% Look through Xtmp names for constant
|
|
% if found, subtract from LHS
|
|
to_remove = [];
|
|
for j = 1:length(Xtmp.name)
|
|
if ~isnan(str2double(Xtmp.name{j}))
|
|
lhssub = lhssub + str2double(Xtmp.name{j}) * Xtmp{j};
|
|
to_remove = [to_remove j];
|
|
end
|
|
% Multiply by -1 now so that it can be added together below
|
|
% Otherwise, it would matter which was encountered first,
|
|
% a parameter on its own or a linear constraint
|
|
Xtmp.(Xtmp.name{j}) = -1 * Xtmp{j};
|
|
end
|
|
for j = length(to_remove):-1:1
|
|
Xtmp = Xtmp.remove(Xtmp.name{j});
|
|
end
|
|
end
|
|
end
|
|
if ~isempty(Xtmp)
|
|
to_remove = [];
|
|
for j = 1:length(Xtmp.name)
|
|
% Handle constraits
|
|
idx = find(strcmp(X.name, Xtmp.name{j}));
|
|
if ~isempty(idx)
|
|
X.(X.name{idx}) = X{idx} + Xtmp{j};
|
|
to_remove = [to_remove j];
|
|
end
|
|
end
|
|
for j = length(to_remove):-1:1
|
|
Xtmp = Xtmp.remove(Xtmp.name{j});
|
|
end
|
|
X = [X Xtmp];
|
|
end
|
|
end
|
|
|
|
clear residual Xtmp plus_node last_node_to_parse to_remove
|
|
|
|
Y = Y - lhssub;
|
|
|
|
%% Check to see if called from Gibbs
|
|
st = dbstack(1);
|
|
varargout = cell(1, 1);
|
|
called_from_olsgibbs = false;
|
|
if strcmp(st(1).name, 'olsgibbs')
|
|
varargout = cell(1, 4);
|
|
called_from_olsgibbs = true;
|
|
end
|
|
|
|
%% Set dates
|
|
fp = max(Y.firstobservedperiod, X.firstobservedperiod);
|
|
lp = min(Y.lastobservedperiod, X.lastobservedperiod);
|
|
if ~isempty(lhssub)
|
|
fp = max(fp, lhssub.firstobservedperiod);
|
|
lp = min(lp, lhssub.lastobservedperiod);
|
|
end
|
|
if isfield(jsonmodel{i}, 'tags') ...
|
|
&& isfield(jsonmodel{i}.tags, 'sample') ...
|
|
&& ~isempty(jsonmodel{i}.tags.sample)
|
|
colon_idx = strfind(jsonmodel{i}.tags.sample, ':');
|
|
fsd = dates(jsonmodel{i}.tags.sample(1:colon_idx-1));
|
|
lsd = dates(jsonmodel{i}.tags.sample(colon_idx+1:end));
|
|
if fp > fsd
|
|
warning(['The sample over which you want to estimate contains NaNs. '...
|
|
'Adjusting estimation range to begin on: ' fp.char])
|
|
else
|
|
fp = fsd;
|
|
end
|
|
if lp < lsd
|
|
warning(['The sample over which you want to estimate contains NaNs. '...
|
|
'Adjusting estimation range to end on: ' lp.char])
|
|
else
|
|
lp = lsd;
|
|
end
|
|
end
|
|
|
|
Y = Y(fp:lp);
|
|
if ~isempty(lhssub)
|
|
lhssub = lhssub(fp:lp);
|
|
end
|
|
pnames = X.name;
|
|
X = X(fp:lp).data;
|
|
|
|
[nobs, nvars] = size(X);
|
|
if called_from_olsgibbs
|
|
varargout{1} = nobs;
|
|
varargout{2} = pnames;
|
|
varargout{3} = X;
|
|
varargout{4} = Y.data;
|
|
varargout{5} = jsonmodel;
|
|
varargout{6} = fp;
|
|
varargout{7} = lp;
|
|
return
|
|
end
|
|
|
|
if isfield(jsonmodel{i}, 'tags') && ...
|
|
isfield(jsonmodel{i}.tags, 'name')
|
|
tag = jsonmodel{i}.tags.('name');
|
|
else
|
|
tag = ['eq_line_no_' num2str(jsonmodel{i}.line)];
|
|
end
|
|
|
|
%% Estimation
|
|
% From LeSage, James P. "Applied Econometrics using MATLAB"
|
|
oo_.ols.(tag).dof = nobs - nvars;
|
|
|
|
% Estimated Parameters
|
|
[q, r] = qr(X, 0);
|
|
xpxi = (r'*r)\eye(nvars);
|
|
oo_.ols.(tag).beta = r\(q'*Y.data);
|
|
oo_.ols.(tag).param_idxs = zeros(length(pnames), 1);
|
|
for j = 1:length(pnames)
|
|
if ~strcmp(pnames{j}, 'intercept')
|
|
oo_.ols.(tag).param_idxs(j) = find(strcmp(M_.param_names, pnames{j}));
|
|
M_.params(oo_.ols.(tag).param_idxs(j)) = oo_.ols.(tag).beta(j);
|
|
end
|
|
end
|
|
|
|
% Yhat
|
|
idx = 0;
|
|
yhatname = [tag '_FIT'];
|
|
if ~isempty(fitted_names_dict)
|
|
idx = strcmp(fitted_names_dict(:,1), tag);
|
|
if any(idx)
|
|
yhatname = fitted_names_dict{idx, 2};
|
|
end
|
|
end
|
|
oo_.ols.(tag).Yhat = dseries(X*oo_.ols.(tag).beta, fp, yhatname);
|
|
|
|
% Residuals
|
|
oo_.ols.(tag).resid = Y - oo_.ols.(tag).Yhat;
|
|
|
|
% Correct Yhat reported back to user
|
|
Y = Y + lhssub;
|
|
oo_.ols.(tag).Yhat = oo_.ols.(tag).Yhat + lhssub;
|
|
|
|
% Apply correcting function for Yhat if it was passed
|
|
if any(idx) ...
|
|
&& length(fitted_names_dict(idx, :)) == 3 ...
|
|
&& ~isempty(fitted_names_dict{idx, 3})
|
|
oo_.ols.(tag).Yhat = ...
|
|
feval(fitted_names_dict{idx, 3}, oo_.ols.(tag).Yhat);
|
|
end
|
|
ds = [ds oo_.ols.(tag).Yhat];
|
|
|
|
%% Calculate statistics
|
|
% Estimate for sigma^2
|
|
SS_res = oo_.ols.(tag).resid.data'*oo_.ols.(tag).resid.data;
|
|
oo_.ols.(tag).s2 = SS_res/oo_.ols.(tag).dof;
|
|
|
|
% R^2
|
|
ym = Y.data - mean(Y);
|
|
SS_tot = ym'*ym;
|
|
oo_.ols.(tag).R2 = 1 - SS_res/SS_tot;
|
|
|
|
% Adjusted R^2
|
|
oo_.ols.(tag).adjR2 = oo_.ols.(tag).R2 - (1 - oo_.ols.(tag).R2)*(nvars-1)/(oo_.ols.(tag).dof);
|
|
|
|
% Durbin-Watson
|
|
ediff = oo_.ols.(tag).resid.data(2:nobs) - oo_.ols.(tag).resid.data(1:nobs-1);
|
|
oo_.ols.(tag).dw = (ediff'*ediff)/SS_res;
|
|
|
|
% Standard Error
|
|
oo_.ols.(tag).stderr = sqrt(oo_.ols.(tag).s2*diag(xpxi));
|
|
|
|
% T-Stat
|
|
oo_.ols.(tag).tstat = oo_.ols.(tag).beta./oo_.ols.(tag).stderr;
|
|
|
|
%% Print Output
|
|
if ~options_.noprint
|
|
if nargin == 3
|
|
title = ['OLS Estimation of equation ''' tag ''' [name = ''' tag ''']'];
|
|
else
|
|
title = ['OLS Estimation of equation ''' tag ''''];
|
|
end
|
|
|
|
preamble = {sprintf('Dependent Variable: %s', jsonmodel{i}.lhs), ...
|
|
sprintf('No. Independent Variables: %d', nvars), ...
|
|
sprintf('Observations: %d from %s to %s\n', nobs, fp.char, lp.char)};
|
|
|
|
afterward = {sprintf('R^2: %f', oo_.ols.(tag).R2), ...
|
|
sprintf('R^2 Adjusted: %f', oo_.ols.(tag).adjR2), ...
|
|
sprintf('s^2: %f', oo_.ols.(tag).s2), ...
|
|
sprintf('Durbin-Watson: %f', oo_.ols.(tag).dw)};
|
|
|
|
dyn_table(title, preamble, afterward, pnames, ...
|
|
{'Estimates','t-statistic','Std. Error'}, 4, ...
|
|
[oo_.ols.(tag).beta oo_.ols.(tag).tstat oo_.ols.(tag).stderr]);
|
|
end
|
|
|
|
if ~called_from_olsgibbs
|
|
varargout{1} = ds;
|
|
end
|
|
end
|
|
end
|
|
|
|
%% Helper Functions
|
|
function ols_error(msg, line)
|
|
error(['ERROR encountered in `dyn_ols` line ' num2str(line) ': ' msg])
|
|
end
|
|
|
|
function [next_plus_node, node_to_parse, last_node_to_parse] = findNextplus_node(plus_node, line)
|
|
% Given an additive entry in the AST, find the next additive entry
|
|
% (next_plus_node). Also find the node that will be parsed into
|
|
% parameter*endogenous||param||exog|endog (node_to_parse).
|
|
% Function used for moving through the AST.
|
|
if ~(strcmp(plus_node.node_type, 'BinaryOpNode') && strcmp(plus_node.op, '+'))
|
|
ols_error('pairs of nodes must be separated additively', line);
|
|
end
|
|
next_plus_node = [];
|
|
last_node_to_parse = [];
|
|
if strcmp(plus_node.arg1.node_type, 'BinaryOpNode') && strcmp(plus_node.arg1.op, '+')
|
|
next_plus_node = plus_node.arg1;
|
|
node_to_parse = getOlsNode(plus_node.arg2, line);
|
|
elseif strcmp(plus_node.arg2.node_type, 'BinaryOpNode') && strcmp(plus_node.arg2.op, '+')
|
|
next_plus_node = plus_node.arg2;
|
|
node_to_parse = getOlsNode(plus_node.arg1, line);
|
|
else
|
|
node_to_parse = getOlsNode(plus_node.arg1, line);
|
|
last_node_to_parse = getOlsNode(plus_node.arg2, line);
|
|
end
|
|
end
|
|
|
|
function node_to_parse = getOlsNode(node, line)
|
|
if ~(strcmp(node.node_type, 'BinaryOpNode') && strcmp(node.op, '*')) ...
|
|
&& ~strcmp(node.node_type, 'VariableNode') ...
|
|
&& ~strcmp(node.node_type, 'UnaryOpNode')
|
|
ols_error('couldn''t find node to parse', line);
|
|
end
|
|
node_to_parse = node;
|
|
end
|
|
|
|
function X = parseTimesNode(ds, node, line)
|
|
% Separate the parameter expression from the endogenous expression
|
|
assert(strcmp(node.node_type, 'BinaryOpNode') && strcmp(node.op, '*'))
|
|
[param, X] = parseTimesNodeHelper(ds, node.arg1, line, {}, dseries());
|
|
[param, X] = parseTimesNodeHelper(ds, node.arg2, line, param, X);
|
|
X = X.rename(param{1});
|
|
for ii = 2:length(param)
|
|
X = [X dseries(X{1}.data, X{1}.firstdate, param{ii})];
|
|
end
|
|
end
|
|
|
|
function [param, X] = parseTimesNodeHelper(ds, node, line, param, X)
|
|
if isOlsParamExpr(node, line)
|
|
param = assignParam(param, node, line);
|
|
elseif isOlsVarExpr(node, line)
|
|
if isempty(X)
|
|
X = evalNode(ds, node, line, X);
|
|
else
|
|
ols_error(['got endog * endog' node.name ' (' node.type ')'], line);
|
|
end
|
|
else
|
|
ols_error('unexpected expression', line);
|
|
end
|
|
end
|
|
|
|
function param = assignParam(param, node, line)
|
|
if ~isempty(param)
|
|
ols_error(['got param * param' node.name ' (' node.type ')'], line);
|
|
end
|
|
param = assignParamHelper(param, node, line);
|
|
end
|
|
|
|
function param = assignParamHelper(param, node, line)
|
|
if strcmp(node.node_type, 'NumConstNode')
|
|
param{end+1} = num2str(node.value);
|
|
elseif strcmp(node.node_type, 'VariableNode')
|
|
param{end+1} = node.name;
|
|
elseif strcmp(node.node_type, 'BinaryOpNode')
|
|
if ~strcmp(node.op, '-')
|
|
ols_error(['got unexpected parameter op ' node.op], line);
|
|
end
|
|
param = assignParamHelper(param, node.arg1, line);
|
|
param = assignParamHelper(param, node.arg2, line);
|
|
else
|
|
ols_error(['got unexpected node (' node.type ')'], line);
|
|
end
|
|
end
|
|
|
|
function tf = isOlsVar(node)
|
|
if strcmp(node.node_type, 'VariableNode') ...
|
|
&& (strcmp(node.type, 'endogenous') ...
|
|
|| (strcmp(node.type, 'exogenous') && any(strcmp(ds.name, node.name))))
|
|
tf = true;
|
|
elseif strcmp(node.node_type, 'UnaryOpNode') ...
|
|
&& (strcmp(node.arg.type, 'endogenous') ...
|
|
|| (strcmp(node.arg.type, 'exogenous') && any(strcmp(ds.name, node.arg.name))))
|
|
tf = true;
|
|
else
|
|
tf = false;
|
|
end
|
|
end
|
|
|
|
function tf = isOlsVarExpr(node, line)
|
|
if strcmp(node.node_type, 'VariableNode') || strcmp(node.node_type, 'UnaryOpNode')
|
|
tf = isOlsVar(node);
|
|
elseif strcmp(node.node_type, 'BinaryOpNode')
|
|
tf = isOlsVarExpr(node.arg1, line) || isOlsVarExpr(node.arg2, line);
|
|
else
|
|
ols_error(['got unexpected type ' node.node_type], line);
|
|
end
|
|
end
|
|
|
|
function X = evalNode(ds, node, line, X)
|
|
if strcmp(node.node_type, 'NumConstNode')
|
|
X = dseries(str2double(node.value), ds.dates, 'const');
|
|
elseif strcmp(node.node_type, 'VariableNode')
|
|
if ~(strcmp(node.type, 'endogenous') ...
|
|
|| (strcmp(node.type, 'exogenous') && any(strcmp(ds.name, node.name))))
|
|
ols_error(['got unexpected type ' node.name ': ' node.type], line);
|
|
end
|
|
X = ds.(node.name)(node.lag);
|
|
elseif strcmp(node.node_type, 'UnaryOpNode')
|
|
Xtmp = evalNode(ds, node.arg, line, X);
|
|
% Only works if dseries supports . notation for unary op (true for log/diff)
|
|
% Otherwise, use: X = eval([node.op '(Xtmp)']);
|
|
X = Xtmp.(node.op);
|
|
elseif strcmp(node.node_type, 'BinaryOpNode')
|
|
Xtmp1 = evalNode(ds, node.arg1, line, X);
|
|
Xtmp2 = evalNode(ds, node.arg2, line, X);
|
|
X = X + eval(['Xtmp1 ' node.op ' Xtmp2']);
|
|
else
|
|
ols_error(['got unexpected node type ' node.node_type], line);
|
|
end
|
|
end
|
|
|
|
function tf = isOlsParam(node)
|
|
if strcmp(node.node_type, 'VariableNode') && strcmp(node.type, 'parameter')
|
|
tf = true;
|
|
else
|
|
tf = false;
|
|
end
|
|
end
|
|
|
|
function tf = isOlsParamExpr(node, line)
|
|
if strcmp(node.node_type, 'NumConstNode')
|
|
tf = true;
|
|
elseif strcmp(node.node_type, 'VariableNode')
|
|
tf = isOlsParam(node);
|
|
elseif strcmp(node.node_type, 'UnaryOpNode')
|
|
tf = false;
|
|
elseif strcmp(node.node_type, 'BinaryOpNode')
|
|
tf = isOlsParamExpr(node.arg1) && isOlsParamExpr(node.arg2);
|
|
if tf && ~strcmp(node.op, '-')
|
|
ols_error(['got unexpected op ' node.op], line);
|
|
end
|
|
else
|
|
ols_error(['got unexpected type ' node.node_type], line);
|
|
end
|
|
end
|