Work in Progress!

Intro

DynaMo (Dynamic Models) is an Open Source package for time series analysis with a special focus on the time series models used in Financial Econometrics.
The package provides methods for simulation, estimation, inference, regularization and prediction of a set of univariate models including:
ARMA, ARMA-GARCH, ACD, MEM.
The package consists of a C library based on the GSL and extensions for popular scientific scripting languages such as R and MATLAB.

Aims

• The R and MATLAB extensions provide a unified package for researchers/end users for the analysis of many time series models customarily used in the financial time series literature.
• The C library provides a unified and flexible structure that allows researchers/developers to extend the package with other model classes and to integrate time series methods in other applications.

Features

• The estimation methods include maximum likelihood as well as penalised maximum likelihood estimation methods.
• The prediction routines allow to automatically obtain forecasts using recursive and rolling forecasting schemes where model parameters are updated each period.

Getting Started in R or MATLAB

# simulate 1000 obs from a garch model with gaussian innovations
y <- rdm( y~garch(1,1), n=1000, param=c(0.2,0.1,0.8))

# estimate
g11 <- dm( y~garch(1,1) )

# show estimation results
summary(g11)

# fitted, actual and residuals plot
plot(g11)

# simulate 1000 obs from a garch model with gaussian innovations
y = rgarch([1 1],1000,[0.2 0.1 0.8]);

# estimate
g11 = garch([1 1],y);

# show estimation results
g11

# fitted, actual and residuals plot
plot(g11)