Last modified: 2007 Aug 16
<< Archive Home Page
Time Series
-
Time series analysis package containin 78 procedures,
including seasonal arima models (also with some parameters set equal to zero),
substet ar models, estimating missing values in time series,
spectrum estimates, and simulation of arma models with normally distributed errors.
(Schlittgen Apr01)
-
linear common trends models of Warne (1993) with analytic impulse responses (Comacho Mar01). There is an associated paper.
-
bqvar.gau performs a Blanchard-Quah decomposition. (Isaac Apr04)
-
johansson.bqvar.src performs a Blanchard-Quah decomposition. (Johansson May04).
Needs johansson.nclag.src
and
johansson.mexp.src.
-
Chow-Lin interpolation procedure (Boldin 1999)
-
Compute grid bootstap confidence intervals for AR process (Hansen 1999 REStat).
(Most recent version.)
- ARCH
-
Estimates a conditionally heteroskedastic and skewed Student T process. (Hansen 1994 IER).
(Most recent version.)
- Polarity Analysis (Chambers Aug00)
- Markov switching (Hamilton 1988-86)
- Vector ARMA Models (Schoenberg Jun00).
Here is the readme file.
- A Markov Chain Monte Carlo (MCMC) Algorithm: Random-Walk Metropolis (Thomas Jun98)
- Fractional differencing operator (Ebens Jun98)
-
Chihwa Kao offers gauss-based econometric package, NPT 1.0, of doing nonstationary panel time series.
The user guide and the program can be downloaded from his website.
-
Panel data procedures (Jönsson 2004):
Start with the documentation and a sample program together with the data files outpg.txt, pgbal.txt, and unemp.txt.
Then get the procedures:
ac.src,
ac_res.src,
est_gls.src,
fe_1w.src,
fe_2w.src,
hac.src,
hausm_m.src,
hsced.src,
hsced_res.src,
hausm_ov.src,
level_fe.src,
lm_test.src,
mixed.src,
regls_1w.src,
hac_2wre.src,
theta_1w.src,
theta_2w.src.
- Hodgson/Vorkink collection of code for
adaptive
estimation. (Linked rather than archived.)
-
Two Hodrick-Prescott Filters
by van Norden, one using banded matrix functions.
(Local copy.)
Additional Hodrick-Prescott Filters
by Matheny, van Norden, and Vigfusson (Apr95), by Lubuele (May97)
by Soderlind (May97).
- Kalman Filter (Nixon93) estimates a time
varying regression model based
on Meinhold and Nozer D. Singpurwalla (1983).
-
Univariate time-series procedures for performing the Beveridge-Nelson
(1981) decomposition: (Jönsson 2004)
infMA.src,
bn.src,
forecast.src,
newbold.src,
cwdecomp.src.
- SS Univariate Time Series (Shuetrim Mar96)
allows full maximum likelihood estimation of
Structural Times Series models in state space form (SSF).
Estimates models that can be
expressed in the SSF, so long as the errors are
Gaussian and the system matrices are time
homogenous.
-
Kalman Filter Gains and Stoffer and Wall bootstrap
has two procedures for users of the TSM module. KF_Gain can
compute Kalman filter gains. SSM_bootstrap is
an implementation of the algorithm of Stoffer and Wall (1991).
-
Four Sample TSM Programs (Roncalli Sep96)
-
Jump And Sharp Cusp Detection By Wavelets (Roncalli Sep96)
-
Time Series Standard Errors (Roncalli Oct96)
computes the standard errors for forecast error
impulse, orthogonalized impulse and forecast error variance
decomposition.
VAR and SVAR
- VAR Routines (Rapach&Wharff Jun96)
for Sim LR lag selection and Runkle's bootstrap.
- VAR Routines (Facklet Mar96)
computes VARS, IRFs, and FEVDs. The code is
modular and allows for alternative identification schemes besides the
usual Cholesky factorization.
-
svar.src (Fackler) automates the estimation of structural VARs.
ARCH and GARCH
- Patton's code, which can be called from inside a program, is a modification of Schoenberg's GARCH code. (Patton Sep99)
- Univariate and multivariate GARCH. (Kroner 1996)
(Also archived at UCSD site, with Fortran code)
Limited documentation included.
Computes impulse response functions and conditional moments tests. The multivariate program estimates the VECH, CCORR, F-ARCH, BEKK and ADC models. Sample data sets and results are provided. This program requires the MAXLIK module from GAUSS.
- t-GARCH (Schoenberg 95)
is a CML program for a t-distributed GARCH(p,q).
- SGARCH.src (Schoenberg Dec95) requires GAUSS 3.2.7+ and the Constrained Maximum Likelihood application module.
- ARCH, GARCH, and IGARCH simulation procedures. (Loretan Aug96)
-
GARCH (Gohs Mar2006) (requires constrained optimization module)
Structural Breaks