def model_training(Y_train, output_path, training_start_date,
training_end_date, chain_len):
X_train = loadX(training_start_date, training_end_date)
X_train = dataFillNA(X_train) # fill na
tmp_columns = X_train.columns.tolist()
tmp_columns.remove('date')
all_data = X_train.merge(Y_train, on='date', how='inner')
X_train = all_data[tmp_columns]
Y_train = all_data['Y']
del all_data
gc.collect()
X_train = Xpoint2Set(X_train, chain_len)
y_train = Ypoint2Set(Y_train, chain_len)
crf = sklearn_crfsuite.CRF(algorithm='lbfgs',
c1=0.1,
c2=0.1,
max_iterations=100,
all_possible_transitions=True)
crf.fit(X_train, y_train)
with open(output_path + 'crf_model.pkl', 'wb') as tmp_fo: # dump model
pickle.dump(crf, tmp_fo)
def model_training(Y_train, output_path, training_start_date,
training_end_date, chain_len):
X_train = loadX(training_start_date, training_end_date)
tmp_columns = X_train.columns.tolist()
tmp_columns.remove('date')
all_data = X_train.merge(Y_train, on='date', how='inner')
cols_to_process = ['OUTSTANDING_CASH_TO_FREE_CAP']
all_data, tot_bin_cols, tot_cutoff_points = getDiscreteFeatures(
all_data, cols_to_process)
tmp_columns = [x for x in tmp_columns if x not in cols_to_process]
tmp_columns = tmp_columns + tot_bin_cols
X_train = all_data[tmp_columns]
Y_train = all_data['Y']
del all_data
gc.collect()
X_train = Xpoint2Set(X_train, chain_len)
y_train = Ypoint2Set(Y_train, chain_len)
crf = sklearn_crfsuite.CRF(algorithm='lbfgs',
c1=0.1,
c2=0.1,
max_iterations=100,
all_possible_transitions=True)
crf.fit(X_train, y_train)
with open(output_path + 'crf_model.pkl', 'wb') as tmp_fo: # dump model
pickle.dump(crf, tmp_fo)
def model_testing(Y_test, output_path, testing_start_date, testing_end_date,
chain_len):
X_test = loadX(testing_start_date, testing_end_date)
X_test = dataFillNA(X_test) # fill na
tmp_columns = X_test.columns.tolist()
tmp_columns.remove('date')
all_data = X_test.merge(Y_test, on='date', how='inner')
X_test = all_data[tmp_columns]
Y_test = all_data['Y']
test_dates = all_data['date']
del all_data
gc.collect()
X_test = Xpoint2Set(X_test, chain_len)
Y_test_pair = Ypoint2Set(Y_test, chain_len)
with open(output_path + 'crf_model.pkl', 'rb') as tmp_fi: # dump model
crf = pickle.load(tmp_fi)
y_pred = crf.predict(X_test)
# test pair
labels = ['-1.0', '1.0']
print(
metrics.flat_classification_report(Y_test_pair,
y_pred,
labels=labels,
digits=3))
# test single
y_pred_single = y_pred[0].copy()
y_pred_single.pop(-1)
y_pred_single.extend([tmp_y[1] for tmp_y in y_pred])
# y_pred_single.insert(0, y_pred[0][0])
y_real_singel = Y_test.astype('str').tolist()
prsc = precision_score(y_real_singel,
y_pred_single,
labels=labels,
average='micro')
print('%s to %s weighted precision: %f' %
(testing_start_date, testing_end_date, prsc))
print('f1 score: %f, precision: %f' %
(metrics.flat_f1_score(
Y_test_pair, y_pred, labels=labels, average='weighted'),
metrics.flat_precision_score(
Y_test_pair, y_pred, labels=labels, average='micro')))
prediction = pd.DataFrame(test_dates)
prediction.loc[:, 'predict'] = y_pred_single
return prediction, prsc
def model_training(y_train, output_path, training_start_date,
training_end_date):
global X_train, Y_train
X_train = loadX(training_start_date, training_end_date)
X_train = dataFillNA(X_train) # fill na
Y_train = y_train
# ==== hyperopt validation
# params = {
# # 'chg_pct': hp.uniform('chg_pct', 0, 0.3),
# # 'chg_threshold': hp.uniform('chg_threshold', 0, 0.3),
# # 'chain_len': hp.randint('chain_len', 9),
# 'training_start_date': training_start_date,
# 'training_end_date': training_end_date
# }
# chg_pct = scipy.stats.uniform(scale=0.3)
# chg_threshold = scipy.stats.uniform(scale=0.3)
# chain_len = scipy.stats.randint(low=2, high =10)
# ==== cross validation
best_cv_score = 0
for tmp_chain_len in range(2, 11): # chain 2~10
# tmp_chg_pct = chg_pct.rvs()
# tmp_chg_threshold = chg_threshold.rvs()
# tmp_chain_len = chain_len.rvs()
# params['chain_len'] = tmp_chain_len
tmp_results = objective(tmp_chain_len)
if tmp_results['cv_score'] > best_cv_score:
best_cv_score = tmp_results['cv_score']
# tmp_sub_params['chg_pct'] = tmp_chg_pct
# tmp_sub_params['chg_threshold'] = tmp_chg_threshold
best_params = tmp_results.copy()
best_params['chain_len'] = tmp_chain_len
# tmp_trial = Trials()
# best_params = fmin(objective, space=params, algo=tpe.suggest, max_evals=100, trials=tmp_trial)
# # get sub-params
# tmp_idx = np.argmin(np.array(tmp_trial.losses()))
# best_params['c1'] = tmp_trial.results[tmp_idx]['c1']
# best_params['c2'] = tmp_trial.results[tmp_idx]['c2']
# best_params['chain_len'] += 2 # adjust chain len
print('best cv score:', best_params['cv_score'])
print('best params:', best_params)
# ==== train with the best params
# Y_train = loadY(best_params['chg_pct'], best_params['chg_threshold'], training_start_date, training_end_date)
# Y_train.loc[:, 'Y'] = Y_train['PeakTrough'].shift(-1) # predict tomorrow !!!!
# Y_train = Y_train.loc[~Y_train['Y'].isnull()] # drop nan
tmp_columns = X_train.columns.tolist()
tmp_columns.remove('date')
all_data = X_train.merge(Y_train, on='date', how='inner')
chain_X_train = all_data[tmp_columns]
chain_Y_train = all_data['Y']
chain_X_train = Xpoint2Set(chain_X_train, best_params['chain_len'])
chain_Y_train = Ypoint2Set(chain_Y_train, best_params['chain_len'])
crf = sklearn_crfsuite.CRF(algorithm='lbfgs',
c1=best_params['c1'],
c2=best_params['c2'],
max_iterations=100,
all_possible_transitions=True)
crf.fit(chain_X_train, chain_Y_train)
# tmp_columns = X_train.columns.tolist()
# tmp_columns.remove('date')
#
# all_data = X_train.merge(Y_train, on='date', how='inner')
# X_train = all_data[tmp_columns]
# Y_train = all_data['Y']
# del all_data
# gc.collect()
#
# X_train = Xpoint2Set(X_train, chain_len)
# Y_train = Ypoint2Set(Y_train, chain_len)
#
# # search parameter by cross validation
# crf = sklearn_crfsuite.CRF(
# algorithm='lbfgs',
# # c1=0.1,
# # c2=0.1,
# max_iterations=100,
# all_possible_transitions=True
# )
#
# params_space = {
# 'c1': scipy.stats.expon(scale=0.5),
# 'c2': scipy.stats.expon(scale=0.05),
# }
#
# labels = ['-1.0', '1.0']
# # val_scorer = make_scorer(precision_score, average='micro', labels=labels)
# val_scorer = make_scorer(metrics.flat_precision_score, average='micro', labels=labels)
#
# rs_cv = RandomizedSearchCV(crf, params_space, cv=3, verbose=10, n_jobs=-1, n_iter=50, scoring=val_scorer) # searching
# rs_cv.fit(X_train, Y_train)
#
# crf = rs_cv.best_estimator_
# # crf.fit(X_train, y_train)
with open(output_path + 'crf_model.pkl', 'wb') as tmp_fo: # dump model
pickle.dump(crf, tmp_fo)
return best_params