Stéphane Adjemia (Scylla)
e6c716ae9b
Added the possibility to use Gauss-Newton in pac/nls.
...
Also added the computation of the covariance matrix of the NLS estimator (using
White and Domovitz approach) and integration test.
2018-11-26 09:53:18 +01:00
Stéphane Adjemia (Scylla)
139c58dd76
Added Gauss Newton algorithm for minimizing sum of squared residuals.
2018-11-26 09:48:50 +01:00
Stéphane Adjemia (Scylla)
cf35496c06
Fixed bug in the evaluation of the sum of squared residuals (PAC/NLS).
2018-11-21 15:21:23 +01:00
Stéphane Adjemia (Scylla)
7d2d0d6590
Added the possibility to estimate the share of optimizing agents by iterative OLS (PAC).
2018-11-21 15:07:15 +01:00
Stéphane Adjemia (Scylla)
dcea7514d9
Merge branch 'master' into ecb-master
2018-10-24 18:31:31 +02:00
Houtan Bastani
8a2c38cf6c
olsgibbs: save fitted values in dataset; allow use of dictionary as in dyn_ols
2018-10-24 17:17:01 +02:00
Stéphane Adjemia (Scylla)
510bc2e299
Removed unused variable.
2018-10-24 12:08:00 +02:00
Stéphane Adjemia (Scylla)
1c1a3ea0a4
Removed tests on targets.
...
A target does not need to be a random walk. Also added a new integration
test (where the target is a constant, 0).
2018-10-24 12:06:14 +02:00
Stéphane Adjemia (Scylla)
a53c63d6d5
Added routines to print PAC_EXPECTATION and VAR_EXPECTATION in files.
...
Each new routine pac.print() and var_expectation.print(), creates two files
that can be included (using the macro language) in a mod file. The first
file, {pac,var}expectationmodename-parameters.inc saved under
{M_.fname}/model/{pac,var}expectationmodel, contains the declaration of the
parameters created for the expanded (var or pac) expectation model, with
calibrated values. The second file, {pac,var}expectationmodename-parameters.inc
saved in the same subfolder, contains the expanded version of the (pac or var)
expectation term as a linear combination of the variables in the companion
representation of the expectatino model.
2018-10-14 16:57:51 +02:00
Stéphane Adjemia (Scylla)
e5317b17de
Cosmetic change.
2018-10-06 17:52:58 +02:00
Stéphane Adjemia (Scylla)
108854a484
Bug fix. verbosity option was not honoured in simplex algorithm (8).
...
Thanks to Reuben for reporting.
2018-10-06 17:50:37 +02:00
Stéphane Adjemia (Scylla)
dfafe86246
Save teh posterior mean of the slope parameters as a column vector.
2018-10-06 16:55:42 +02:00
Stéphane Adjemian (Charybdis)
82187f9d41
Fixed olsgibbs routine.
2018-10-06 16:08:28 +02:00
Stéphane Adjemian (Charybdis)
798aaaa968
Cosmetic change.
2018-10-04 23:17:49 +02:00
Houtan Bastani
ee2ad1ee15
olsgibbs: linear regression with Gibbs sampler
2018-10-04 17:12:47 +02:00
Houtan Bastani
57bf291c00
add comment
2018-10-04 17:12:47 +02:00
Stéphane Adjemian(Charybdis)
ae71599292
Switch to new get_companion_matrix routine.
2018-09-28 19:27:17 +02:00
Stéphane Adjemian(Charybdis)
bbbaabc702
Cosmetic change.
2018-09-28 12:26:59 +02:00
Michel Juillard
3ad654b9b8
Allowing for different number of periods in conditional_forecast_path.
...
Updated preprocessor submodule.
2018-09-27 18:58:24 +02:00
Stéphane Adjemian(Charybdis)
07141a8681
Various optimizations.
2018-09-26 22:51:05 +02:00
Stéphane Adjemian(Charybdis)
7c83ba7ea7
Fixed documentation header.
2018-09-26 22:51:05 +02:00
Houtan Bastani
973c3955f0
adjust for ‘targets’ keyword instead of ‘trends’
2018-09-25 14:58:36 +02:00
Houtan Bastani
974bcd92ae
Merge branch 'ecb-master' into ecb-master-targets
2018-09-25 14:41:11 +02:00
Houtan Bastani
c6d5b09c29
fix typo
2018-09-25 14:39:48 +02:00
Stéphane Adjemian(Charybdis)
b3a8062b46
Added missing output.
2018-09-19 17:39:58 +02:00
Stéphane Adjemian(Charybdis)
031e2c87c6
New version of get_companion_matrix + new integration test.
...
The routine is still buggy. In PAC and VAR_EXPECTATION models we
use get_companion_matrix_legacy routine instead.
2018-09-19 17:39:48 +02:00
Stéphane Adjemian(Charybdis)
2cf6ccdaa5
Updated dseries submodule.
2018-09-18 17:37:37 +02:00
Stéphane Adjemian(Charybdis)
3e6e448134
Updated dates and dseries submodules.
2018-09-14 23:13:13 +02:00
Stéphane Adjemian(Charybdis)
99aadf7d1a
Updated m-unit-tests submodule.
2018-09-14 15:37:30 +02:00
Houtan Bastani
62301fa1e9
dyn_ols: fix typo
2018-09-14 15:09:31 +02:00
Stéphane Adjemian(Charybdis)
2518cd9884
Raise error of an optimization algorithm is not available.
2018-09-13 16:16:07 +02:00
Stéphane Adjemian(Charybdis)
2437d7034e
Give access to more than one optimizer in PAC estimation (NLS).
...
Default is to use csminwel optimizer (because it seems to do the job and does
not rely on the Mathworks toolbox).
2018-09-13 16:15:52 +02:00
Houtan Bastani
eb316d7e2d
update files for change of option name from `trends` to `targets`
2018-09-13 12:24:32 +02:00
Houtan Bastani
5ab8120780
get companion matrix using preprocessor output instead of get_ar_ec_matrices.m
2018-09-12 18:38:26 +02:00
Houtan Bastani
74094683de
get_companion_matrix: fix header
2018-09-12 18:29:21 +02:00
Houtan Bastani
a6150bae82
get_companion_matrix: simplify check (no need for allocation)
2018-09-12 18:29:14 +02:00
Houtan Bastani
a32e2f165a
get_companion_matrix: remove unnecessary conditional
2018-09-12 18:29:07 +02:00
Houtan Bastani
e9d7f6ca24
Merge branch 'master' into ecb-master
2018-09-12 18:28:47 +02:00
Stéphane Adjemian(Charybdis)
6815624a6a
Fixed bug.
...
Growth neutrality correction was applied twice.
2018-09-12 15:41:07 +02:00
Sébastien Villemot
51416bf2ae
Always use "isoctave" for testing Octave context
2018-09-07 17:22:08 +02:00
Stéphane Adjemian(Charybdis)
0f3678ec4a
Fixed ordering of variables in error correction term of PAC equation
...
PAC equation has to be written as
diff(x) = a0*(xstar(-1)-x(-1)) + a1*diff(x(-1)) + ... + ap*diff(x(-p)) + PAC_EXPECTATION(pacmodelname) + ...;
In the error correction term, a0*(xstar(-1)-x(-1)), we must have the difference
between the target (the trend xstar(-1)) and the level of the endogenous
variable (x(-1)). To ensure stability around the trend, the parameter a0 needs
to be positive.
REMARKS
[1] In the TREND_COMPONENT_MODEL the error correction terms are written in
reverse order, ie as the difference betwwen the level of the endogenous
variable and the trend variable.
[2] In the estimation routine we do not constrain a0 to be positive, but is
would surely help to satisfy this condition in the initial condition.
2018-09-06 22:07:31 +02:00
Stéphane Adjemian(Charybdis)
74ef4d2131
Cosmetic change.
2018-09-06 13:40:43 +02:00
Stéphane Adjemian(Charybdis)
5d8a1feda4
Fixed regular expression.
...
Data matrices were not correcty lagged for lags>1. This commit fixes the
inconsistencies (in terms of residuals) between Iterative OLS and NLS routines.
2018-09-06 13:40:27 +02:00
Stéphane Adjemian(Charybdis)
940c970ff3
Use same range in NLS and Iterative OLS.
...
The range option is the range of data considered for the right hand side variable.
2018-09-04 09:40:47 +02:00
Stéphane Adjemian(Charybdis)
6e1d06bddc
Display SSR in Iterative OLS.
2018-09-03 18:01:59 +02:00
Stéphane Adjemian(Charybdis)
8b1b1273c3
Fixed correction for growth neutrality.
...
Also added the possibility to use an endogenous variable for growth.
2018-09-03 17:57:10 +02:00
Stéphane Adjemian(Charybdis)
fde103d5f4
Removed debugging display.
2018-09-03 17:55:55 +02:00
Stéphane Adjemian(Charybdis)
a207550680
Fixed error correction term.
2018-09-03 17:53:01 +02:00
Stéphane Adjemian(Charybdis)
044f0f8de6
Added missing output required by some integration tests.
2018-09-03 17:38:09 +02:00
Stéphane Adjemian(Charybdis)
5ea0c8098b
Ensure that the ordering of the trend variables is consistent...
...
... With the ordering of the EC variables.
2018-09-03 17:35:59 +02:00