**Code for the space-time ETAS model and
stochastic declustering **

Source code and
example dataset:

System requirement:

OS:
Linux

Software: gfortran (or
Intel fortran), mpich2

The code here is in a
parallel version. Suppose that you have already installed mpich2. (Not yet?! Please go to
http://www.mcs.anl.gov/research/projects/mpich2/ to have a look.)

(If
you need a non-parallel version, please use this
file. This program is coded by Dr. Qi Wang by modifying the parallel
version. However, I cannot provide much technical support for this serial
version)

After downloading sd.tar.gz,
please uncompress the file in a fold by

*$ tar -zxvf sd.tar.gz
*

and compile it by

*$ mpif77 *.f -o
etas8p*

and run the example dataset by

*$ export
GFORTRAN_UNBUFFERED_ALL='y'*

*$ mpiexec -n 8 etas8p < jap.in *

There is 1 example
dataset:

jap.in: answers to the input
prompts.

alljpM45.etas: JMA
catalogue of shallow eqs. M>=4.5.

jap.out: outputs to stdout.

Note: the polygon must
be input in the counterclockwise order.

The outputs of the
program will be

para: MLE of paramters for
each step (mu, A, c, alpha, p, D,
q,

gamma).

rates.dat: first row: m n

first column of Row 2 to Row last: m x n matrix
of total rate

second column of Row 2 to Row Last: m x n
matrix of

background rate divided by mu

third column of Row 2 to Row Last: m x n matrix
of

intensity rate at the end of time span of
dataset.

probs.dat: first column : EQ sequence n

second column: background probability for this
event.

third column: background rate

fourth column: bandwidth

pmatr.dat : each row: j, i, rho_{ij} (probabability that j is trigger by i)

!Note: It is not
warranted that you can get readable results with all earthquake catalogs. By my
experience, the catalog must be complete above the magnitude threshold in the
target region, and there should be no increasing trend in the catalog.

If you find this
program useful in your research, please refer to the following articles in your
publications.

1.
__Zhuang____ J.__ (2006) *Second-order
residual analysis of spatiotemporal point processes and applications in model
evaluation.* **Journal of the Royal Statistical Society: Series B
(Statistical Methodology)**, 68 (4), 635-653. doi: 10.1111/j.1467-9868.2006.00559.x.

2.
__Zhuang____ J.__, Ogata Y. and Vere-Jones D. (2004). *Analyzing
earthquake clustering features by using stochastic reconstruction*. **Journal of Geophysical Research**, 109, No. B5, B05301,
doi:10.1029/2003JB002879.

3.
__Zhuang____ J.__, Ogata Y. and Vere-Jones D. (2002). *Stochastic declustering of space-time earthquake occurrences*.
**Journal of the American Statistical Association**, 97: 369-380.

*updated** on 24/April/2011*.