# step 3.3 - predict subject and join it to the readings data
print('Predicting subject ... ')
predicted_subject = clf_subject.predict(readings_test)
readings_subject_test = np.column_stack((readings_test, predicted_subject))
# step 3.4 - predict activity
print('Predicting activity ... ')
predicted_activity = clf_activity.predict(readings_subject_test)
# step 4 - printing results
actual_activity_test = df_test_s.ix[:, -1]
actual_subject_test = df_test_s.ix[:, -2]
subj_activity_test = np.array((100 * actual_subject_test) +
actual_activity_test)
predicted_subj_activity_test = (100 * predicted_subject) + predicted_activity
subj_activity_train = (100 * subj_train_for_result) + activity_train_for_result
predicted_subj_activity_train = (
100 * predicted_subj_train) + predicted_activity_train
ResultsWriter.write_to_file('results_junquan.txt',
model='gnb_pca',
y_train_actual=subj_activity_train,
y_train_predicted=predicted_subj_activity_train,
y_test_actual=subj_activity_test,
y_test_predicted=predicted_subj_activity_test,
dur_train_activity=dur_train_activity,
dur_train_subj=dur_train_subj,
dur_train_both=0,
method='sa') # method = both / as / sa
})
# step 2.2 - predict subject activity
print('Predicting subject activity ... ')
predicted_subj_activity = clf_multi.predict(readings_test)
predicted_subj_activity = pd.DataFrame({
'subject':
predicted_subj_activity[:, 1],
'activity_id':
predicted_subj_activity[:, 0]
})
predicted_subj = predicted_subj_activity.ix[:, 1]
predicted_activity = predicted_subj_activity.ix[:, 0]
predicted_subj_activity_test = (100 * predicted_subj) + predicted_activity
# step 3 - printing results
actual_subj = df_test.ix[:, -3]
actual_activity = df_test.ix[:, -2]
subj_activity_test = (100 * actual_subj) + actual_activity
subj_activity_train = (100 * subj_train) + activity_train
ResultsWriter.write_to_file('results_junquan.txt',
model='gnb_multioutput',
y_train_actual=subj_activity_train,
y_train_predicted=predicted_subj_activity_train,
y_test_actual=subj_activity_test,
y_test_predicted=predicted_subj_activity_test,
dur_train_activity=0,
dur_train_subj=0,
dur_train_both=dur_train_both,
method='both') # method = both / as / sa
# step 1.2 - fit the model to predict subject
print('Fitting model to predict subject and activity...')
clf_both = SGDClassifier(alpha=0.1)
time_bgn = time.time()
clf_both.fit(readings_train, subj_activity_train)
dur_train_both = time.time() - time_bgn
predicted_subj_activity_train = clf_both.predict(readings_train)
# step 2.1 - get the readings data (from data stratified using subject)
print('Predicting subject activity ... ')
readings_test = df_test.ix[:, :-3]
# step 2.2 - predict subject activity
predicted_subj_activity_test = clf_both.predict(readings_test)
# step 3 - printing results
subj_activity_test = df_test.ix[:, -1]
ResultsWriter.write_to_file(
'results_junquan_both.txt',
model='svm_sgd_run_' + str(i + 1),
y_train_actual=subj_activity_train,
y_train_predicted=predicted_subj_activity_train,
y_test_actual=subj_activity_test,
y_test_predicted=predicted_subj_activity_test,
dur_train_activity=0,
dur_train_subj=0,
dur_train_both=dur_train_both,
method='both') # method = both / as / sa