def prepare_data(analysis):
channel_dir, info_dir, global_settings = ut.find_settings()
scenario = global_settings['scenario']
reader = hpr.HHParameterReader(channel_dir, scenario)
preferences = reader.parameters
preferences['trainvars'] = preferences['all_trainvar_info'].keys()
startTime = datetime.now()
print('data loading is started: ' + str(startTime))
if analysis == 'HHmultilepton':
normalizer = hht.HHDataNormalizer
loader = hht.HHDataLoader(normalizer, preferences, global_settings)
elif analysis == 'HHbbWW':
normalizer = bbwwt.bbWWDataNormalizer
loader = bbwwt.bbWWLoader(normalizer, preferences, global_settings)
data = loader.data
print('data loading is finished')
print(datetime.now() - startTime)
scenario = global_settings['scenario']
scenario = scenario if 'nonres' in scenario else 'res/' + scenario
hyperparameters_file = os.path.join(
os.path.expandvars('$CMSSW_BASE'),
'src/machineLearning/machineLearning/info/',
global_settings['process'], global_settings['channel'], scenario,
'hyperparameters.json')
with open(hyperparameters_file, 'rt') as in_file:
preferences['hyperparameters'] = json.load(in_file)
return data, preferences, global_settings
def main(to_continue, opt_dir, bbww):
if not to_continue:
settings_dir = os.path.join(
os.path.expandvars('$CMSSW_BASE'),
'src/machineLearning/machineLearning/settings')
global_settings = ut.read_settings(settings_dir, 'global')
output_dir = os.path.expandvars(global_settings['output_dir'])
else:
settings_dir = os.path.join(opt_dir, 'run_settings')
global_settings = ut.read_settings(settings_dir, 'global')
output_dir = opt_dir
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
if not to_continue:
if not os.path.exists(os.path.join(output_dir, 'run_settings')):
ut.save_run_settings(output_dir)
if not os.path.exists(os.path.join(output_dir, 'run_info')):
ut.save_info_dir(output_dir)
# use_scratch_for_data(global_settings)
print("::::::: Reading parameters :::::::")
if global_settings['ml_method'] == 'xgb':
param_file = os.path.join(settings_dir, 'xgb_parameters.json')
else:
param_file = os.path.join(settings_dir, 'nn_parameters.json')
hyperparameter_info = ut.read_json_cfg(param_file)
pso_settings = ut.read_settings(settings_dir, 'pso')
pso_settings.update(global_settings)
addition = ut.create_infoPath_addition(global_settings)
channel_dir = os.path.join(output_dir, 'run_info')
info_dir = os.path.join(channel_dir, addition)
scenario = global_settings['scenario']
reader = hpr.HHParameterReader(channel_dir, scenario)
preferences = reader.parameters
normalizer = hht.HHDataNormalizer if not bbww else bbwwt.bbWWDataNormalizer
if os.path.exists(preferences['data_csv']):
print(':::::::: Loading data from .csv file ::::::::')
data = pandas.read_csv(preferences['data_csv'])
else:
print('::::::: Loading data to be saved to pandas.DataFrame :::::::')
if not bbww:
loader = hht.HHDataLoader(normalizer, preferences, global_settings)
else:
loader = bbwwt.bbWWLoader(normalizer, preferences, global_settings)
data = loader.data
if global_settings['ml_method'] in ['lbn', 'nn']:
use_Wjets = True
if 'bb2l' in global_settings['channel']:
use_Wjets = False
data = mt.multiclass_encoding(data, use_Wjets)
loader.save_to_csv()
print("\n============ Starting hyperparameter optimization ==========\n")
swarm = pt.ParticleSwarm(pso_settings, st.get_fitness_score,
hyperparameter_info, to_continue, output_dir)
optimal_hyperparameters = swarm.particleSwarmOptimization()[0]
print("\n============ Saving results ================\n")
best_parameters_path = os.path.join(output_dir,
'best_hyperparameters.json')
ut.save_dict_to_json(optimal_hyperparameters, best_parameters_path)
print("Results saved to " + str(output_dir))
def split_data(global_settings, preferences):
print('============ Starting evaluation ============')
if os.path.exists(preferences['data_csv']):
data = pandas.read_csv(preferences['data_csv'])
else:
normalizer = bbwwt.bbWWDataNormalizer
loader = bbwwt.bbWWLoader(normalizer, preferences, global_settings)
data = loader.data
hhvt.plot_trainvar_multi_distributions(data, preferences['trainvars'],
global_settings['output_dir'])
hhvt.plot_correlations(data, preferences['trainvars'], global_settings)
keysNotToSplit = []
if '3l_1tau' in global_settings['channel']:
keysNotToSplit = ['WZ', 'DY', 'TTTo']
print('These keys are excluded from splitting: ', keysNotToSplit)
evtNotToSplit = (data['key'].isin(keysNotToSplit))
evtEven = (data['event'].values % 2 == 0)
evtOdd = ~(data['event'].values % 2 == 0)
even_data = data.loc[np.logical_or(evtEven, evtNotToSplit)]
odd_data = data.loc[np.logical_or(evtOdd, evtNotToSplit)]
return even_data, odd_data
def create_data_dict(preferences, global_settings):
normalizer = bbwwt.bbWWDataNormalizer
loader = bbwwt.bbWWLoader(normalizer, preferences, global_settings)
data = loader.data
hhvt.plot_single_mode_correlation(data, preferences['trainvars'],
global_settings['output_dir'],
'trainvar')
hhvt.plot_trainvar_multi_distributions(data, preferences['trainvars'],
global_settings['output_dir'])
sumall = data.loc[data["process"] == "TT"]["totalWeight"].sum() \
+ data.loc[data["process"] == "W"]["totalWeight"].sum() \
+ data.loc[data["process"] == "DY"]["totalWeight"].sum() \
+ data.loc[data["process"] == "ST"]["totalWeight"].sum() \
+ data.loc[data["process"] == "Other"]["totalWeight"].sum() \
+ data.loc[data["target"] == 1]["totalWeight"].sum()
print(
"TT:W:DY:ST:Other:HH \t" \
+ str("%0.3f" %(data.loc[data["process"] == "TT"]["totalWeight"].sum()/sumall)) \
+ ":" + str("%0.3f" %(data.loc[data["process"] == "W"]["totalWeight"].sum()/sumall)) \
+ ":" + str("%0.3f" %(data.loc[data["process"] == "DY"]["totalWeight"].sum()/sumall)) \
+ ":" + str("%0.3f" %(data.loc[data["process"] == 'ST']["totalWeight"].sum()/sumall)) \
+ ":" + str("%0.3f" %(data.loc[data["process"] == 'Other']["totalWeight"].sum()/sumall)) \
+ ":" + str("%0.3f" %(data.loc[data["target"] == 1]["totalWeight"].sum()/sumall))
)
use_Wjet = True
if 'bb2l' in global_settings['channel']:
use_Wjet = False
data = mt.multiclass_encoding(data, use_Wjet)
hhvt.plot_correlations(data, preferences["trainvars"], global_settings)
even_data = data.loc[(data['event'].values % 2 == 0)]
odd_data = data.loc[~(data['event'].values % 2 == 0)]
data_dict = {
'trainvars': preferences['trainvars'],
'odd_data': odd_data,
'even_data': even_data
}
return data_dict