def risk_nutrition(id_user, comp=False, db=None):
"""
Give a risk using nutrition information
WARNING: untested rules
"""
if db is None:
db = create_database_connection()
score = 0
if not comp:
return 0
# obtain the responses
ans = db.get_responses_category(id_user=id_user, phase=2)
# load the rules file
table = read_csv('food_model.csv', sep=';')
for _, row in table.iterrows():
# Item Table Group
table_ = round(row['Table'])
answer = ans[row['Item']]
if row['Daily'].strip().lower() == 'yes':
answer *= 7
if table_ == 1:
score += table_1(group=row['Group'], n=answer)
elif table_ == 2:
score += table_2(answer)
elif table_ == 3:
score += table_3(answer)
return score * 10 / table.shape[0]
def risk_activity(id_user, comp=False, db=None):
if db is None:
db = create_database_connection()
if not comp:
return 0
ans = db.get_responses_category(id_user=id_user, phase=3)
# compute the METS-min/week
METS = ans[0] * ans[1] * 8 + ans[2] * ans[3] * 4 + ans[4] * ans[5] * 3.3
# this have to be reviewed for sure!
superior_limit = 1800
return (min(METS, superior_limit) / superior_limit) * 100
def obesity_risk(id_user, completed=None, network=True):
"""
Update for "explained version"
:param network:
:param id_user:
:param completed:
:return: score, dict with the subscores
"""
# make connection
db = create_database_connection()
if completed is None:
completed = db.check_completed(id_user)
# coeficient for each part
coef = (0.33, 0.34, 0.33)
bmi_score, bmi = risk_bmi(id_user, db)
part_1 = bmi_score * coef[0]
part_2 = risk_nutrition(id_user, completed[1], db) * coef[1]
part_3 = risk_activity(id_user, completed[2], db) * coef[2]
network_correction, n_contacts, mean_contacts = 0, 0, 0
raw_wakaestado = min(part_1 + part_2 + part_3, 100)
if network:
network_correction, n_contacts, mean_contacts = network_influence(
id_user, raw_wakaestado, db, all(completed))
final_wakaestado = min(raw_wakaestado + network_correction, 100)
# pack the different parts
partial_scores = {
'wakascore': final_wakaestado,
'bmi_score': bmi_score,
'nutrition': part_2 / coef[1],
'activity': part_3 / coef[2],
'network': network_correction * 10, # visualizacion
'bmi': bmi,
'n_contacts': n_contacts,
'mean_contacts': ceil(mean_contacts)
}
# store the last wakaestado
db.set_last_wakaestado(id_user, final_wakaestado)
# close stuff
db.close()
del db
return final_wakaestado, partial_scores
def risk_bmi(id_user, db=None):
"""
Gives a risk score
this is a modular function in order be easier to update
"""
if db is None:
db = create_database_connection()
bmi = db.getBMI(id_user)
if bmi == 0: # sanity check
return 0, 0
if bmi >= 40:
return 0, bmi
elif 35 <= bmi < 40:
return 25, bmi
elif 30 <= bmi < 35:
return 50, bmi
elif 25 <= bmi < 30:
return 75, bmi
elif bmi < 25:
return 100, bmi
time.sleep(float(environ.get('CONSULTING_TIME', 0.4)))
else:
# if no messages lets be *more* gentle with telegram servers
time.sleep(float(environ.get('NO_MESSAGE_TIME', 0.8)))
except Exception as e:
logger.error(e)
logger.error('Bot to sleep!')
time.sleep(float(environ.get('ERROR_TIME', 2)))
if __name__ == '__main__':
# TODO QUESTIONS ON CACHE FOR NEXT VERSION
init_date = datetime.datetime.now()
db = create_database_connection()
db.setup()
# caching the number of questions
nq_category = db.n_questions()
# very important, this password is used to serve the data as file
password_data = environ.get('PASSWORD_DATA')
# default language
def_lang_ = environ.get('DEFAULT_LANG', 'es')
# link to the network
network_link = environ.get('NETWORK_LINK').replace('_', '\\_')
network_filename = environ.get('NETWORK_FILENAME', 'netweb').lower()
# god mode
god_mode = environ.get('GOD_MODE').lower()