The preferred method for installing HORNET is to create an isolated installation location, using virtualenv, at HORNET_HOME/bootstrap.
$ paver bootstrap
$ source bootstrap/bin/activate
(bootstrap)$ paver install
(bootstrap)$ hornet --help
Usage: hornet [options]
Options:
--version show program's version number and exit
-h, --help show this help message and exit
-a ACTION, --action=ACTION
Perform the ACTION (one of run, profile, test,
coverage)
-c CONFIG_FILE, --config=CONFIG_FILE
Use the configuration in CONFIG_FILE (defaults
to './config_hornet.py')
-p, --psyco Turn on psyco compilation
Hornet is configured in config_hornet.py using a hornet.application.Config object. A sample config_hornet.py:
import hornet
import hornet.plugin.networkbuilder
import hornet.plugin.abstractor
import hornet.plugin.ruleminer
import hornet.db
from datetime import date, timedelta
c = hornet.Config()
c.output_dir = 'results'
# Set up periods
c.start = date(2006, 1, 1)
c.end = date(2006, 5, 28)
c.period_size = timedelta(7)
db = hornet.db.OracleDb('scott', 'tiger', 'localhost:1521/orcl')
# Network Builder
c.plugins['builder'] = hornet.plugin.networkbuilder.NetworkBuilder()
c.plugins['builder'].builder = hornet.plugin.networkbuilder.DatabaseGraphBuilder
c.plugins['builder'].builder.db = db
# Abstract graph
c.plugins['abstract'] = hornet.plugin.abstractor.GraphAbstraction()
c.plugins['abstract'].provider = hornet.plugin.abstractor.DatabaseAttributeProvider(db)
c.plugins['abstract'].listen_to = [c.plugins['builder']]
# Rule Decay
c.plugins['ruleslen'] = hornet.plugin.ruleminer.BaseMiner()
c.plugins['ruleslen'].listen_to = [c.plugins['builder']]
# Display Association Rules and Rule Decay
c.plugins['rules'] = hornet.plugin.ruleminer.RuleMiner()
c.plugins['rules'].listen_to = [c.plugins['abstract']]