def diagnose_condition(symptoms, age_group=None):
"""Prevents the segfault from bringing the entire process down.
This function *MUST* be run in a background thread.
For example:
symptoms = ['fever', 'swollenglands']
t = threading.Thread(target=diagnose_condition, args=(symptoms))
t.start()
t.join()
print diagnosis
>>> 'mumps'
"""
global diagnosis
from doctor import Doctor
d = Doctor('medical.pl')
diagnosis = d.diagnose(symptoms, age_group)
return
def diagnose_condition(symptoms, age_group=None):
"""Prevents the segfault from bringing the entire process down.
This function *MUST* be run in a background thread.
For example:
symptoms = ['fever', 'swollenglands']
t = threading.Thread(target=diagnose_condition, args=(symptoms))
t.start()
t.join()
print diagnosis
>>> 'mumps'
"""
global diagnosis
from doctor import Doctor
d = Doctor('medical.pl')
diagnosis = d.diagnose(symptoms,age_group)
return
df = pd.read_csv('./Training.csv') # training data df
df_test = pd.read_csv('./Testing.csv') # testing data df
#print(df)
#print(df_test)
# encode prognosis values and store them
le = preprocessing.LabelEncoder()
le.fit(pd.concat([df['prognosis'], df_test['prognosis']]))
encoded_prognosis = le.transform(df['prognosis'])
# initialize and train doctor
doctor = Doctor()
doctor.train(df)
# get diagnosis with testing data
y_diagnosis = doctor.diagnose(df_test[df_test.columns.difference(['prognosis'
])])
y_pred = [
diagnosis_entry["diagnosis_code"] for diagnosis_entry in y_diagnosis
] # predicted encoded diagnosis values
y_proba = [diagnosis_entry["probability"] for diagnosis_entry in y_diagnosis
] # probability vakues for each preciction
# compare your prediction with actual values
y_true = doctor.diagnosis_encoder.transform(df_test['prognosis'])
print("Accuracy:", metrics.accuracy_score(y_true, y_pred))
print("Classification Report:")
print(classification_report(y_true, y_pred, target_names=df_test['prognosis']))
print("\nTesting Data Results")
test_res = []
