def calculate_probability_for_disease(country, user_symptom_list):
"""Returns diseases based on symptoms the user has"""
disease_probabilty = []
diseases_in_country = data.disease_prob_for_country[country]
normalizing_term = 0
for disease in diseases_in_country:
prob = 0
if len(user_symptom_list) == 0:
prob = helpers.get_probability_for_disease(disease['disease'],country)
else:
for symptom in user_symptom_list:
if symptom in data.symptoms_for_disease[disease['disease']]:
prob += disease["probability"] * helpers.get_symptom_probability(symptom)
normalizing_term += prob
disease_probabilty.append({ 'disease': disease["disease"], 'probability': prob })
for disease in disease_probabilty:
disease["probability"] /= normalizing_term
return disease_probabilty
def get_ordered_symptom_list(country, user_symptom_list, symptom_blacklist):
"""Returns symptoms to ask the user about based on existing symptoms and location"""
diseases = data.disease_prob_for_country[country]
disease_probabilities = {}
tmp = calculate_probability_for_disease(country, user_symptom_list)
for d in tmp:
disease_probabilities[d['disease']] = d['probability']
symptoms = {}
for disease in diseases:
tmp = data.symptoms_for_disease[disease['disease']]
for symptom in tmp:
if symptom in user_symptom_list:
continue
if symptom in symptom_blacklist:
continue
sprob = helpers.get_symptom_probability(symptom)
sprob *= disease_probabilities[disease['disease']]
if symptom not in symptoms:
symptoms[symptom] = sprob
else:
symptoms[symptom] += sprob
return sorted([{'symptom': z[0], 'prob': z[1]} for z in symptoms.iteritems()], reverse=True, cmp=lambda x,y: cmp(x['prob'], y['prob']))