def main(argv):
l_config = renter_constants.learned_config
if len(argv) == 2:
# - Convert the raw data to SETI.
which = argv[1]
if which == '1':
l_config = renter_constants.learned_config2
elif which == '2':
l_config = renter_constants.learned_config2
l_config.raw_filenames = ['data/from_haoran/clean_*.csv']
elif which == '3' or which =='regression_test':
l_config = renter_constants.learned_config2
l_config.raw_filenames = ['data/new_data_set/clean_full_all.csv'] #['data/round1/prices_samples_full_0921212303_all.csv']
if which == 'regression_test':
model_cfg.change_dirs('tmp/regression_test', l_config.model_configs)
setis = logs_to_seti.generate_seti(l_config.raw_filenames)
#setis = setis[0:1000]
print 'Generated: %d setis' % len(setis)
if _DEBUG:
f = open('tmp/cur_run_setis.csv', 'wb')
setis_txt = '\n'.join([str(seti) for seti in setis])
f.write(setis_txt)
f.close()
run_pipeline.run(l_config.model_configs, setis)
return {
'setis': setis,
'l_config': l_config
}
def testLogsToSeti():
# First three lines are copied from tdg_v0.csv
csvs = ['testdata/logs_to_seti_csv0.csv']
# s0. dob: 05/06/1989. gender:f
# s1. dob: 05/06/1949. gender:m
s0 = seti.create_seti(19.15, bfs=[('has_bite_dog', 'N')], cfs=[('dob', 26.0)])
s1 = seti.create_seti(22.51, bfs=[('has_bite_dog', 'N')], cfs=[('dob', 66.0)])
# Test model gets created and loaded.
# Test that we can score one example.
setis = logs_to_seti.generate_seti(csvs, for_test=True)
wants = [s0, s1]
for i in xrange(len(setis)):
assertEquals(str(wants[i]), str(setis[i]))
