ADE/app.py at master · jsleeo424/ADE · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
import streamlit as st
import numpy as np
import pandas as pd
import pickle
from sklearn.ensemble import RandomForestClassifier
from random import randrange

@st.cache(suppress_st_warning=True, allow_output_mutation=True)
def load_pretrained_model(path):
    loaded_model = pickle.load(open(path, 'rb'))
    return loaded_model

def inference_model(model, data):
    """
    Here, we assume that the data
    """
    confidence = np.amax(model.predict_proba(data.reshape(1, -1)))
    decision = model.predict(data.reshape(1, -1))
    return confidence, decision

st.title("ADE: ADHD prediction from EveryDay features")
st.write("To use ADE, please fill the form below.")
with st.form("ADHD_form"):
    st.header("About the subject ")
    height_inch = st.number_input('Your Height (inches)')
    weight_pound = st.number_input('Your Weight (pounds)')

    sleep_hours = st.slider('How many hours do you sleep per day? (hours/day)', 0.0, 24.0)
    income_desc = st.radio(
        'What is your TOTAL COMBINED FAMILY INCOME for the past 12 months? ',
        ('Less than $5,000', '$5,000 through $11,999', '$12,000 through $15,999',
        '$16,000 through $24,999', '$25,000 through $34,999', '$35,000 through $49,999',
        '$50,000 through $74,999', '$75,000 through $99,999', '$100,000 through $199,999',
        '$200,000 and greater'))

    st.subheader("Screen time")
    st.text("Please fill the screen time during typical weekdays and weekends.")
    st.text("Here, the screen time is the sum of ")

    st.markdown(' - Watch videos (such as YouTube)')
    st.markdown(' - Play video games on a computer, console, phone or other device')
    st.markdown(' - Text on a cell phone, tablet, or computer')
    st.markdown(' - Visit social networking sites like Facebook, Twitter, Instagram')
    st.markdown(' - Video chat (Skype, Facetime, etc.)')

    screen_week = st.slider('Screen time during typical weekdays (hours/day)', 0.0, 24.0)

    screen_weekend = st.slider('Screen time during typical weekends (hours/day)', 0.0, 24.0)

    st.header("About the biological mother of the subject")
    drug_during_pregnancy = st.multiselect('Have you ever taken substances below during pregancy?',
                                    ['Tobacco', 'Alcohol', 'Marijuana', 'Cocaine/Crack',
                                    'Heroin/Morphine', 'Oxycontin', 'Any other drugs', 'None'])

    submitted = st.form_submit_button("Submit")
    if submitted:
        st.success("Successfully submitted the form.")
        subject_bmi = (weight_pound * 703) / ((height_inch)**2)
        #st.write('BMI of the subject is ', subject_bmi)

        income_category_num = 0
        if income_desc == "Less than $5,000":
            income_category_num=1
        elif income_desc == "$5,000 through $11,999":
            income_category_num=2
        elif income_desc == "$12,000 through $15,999":
            income_category_num=3
        elif income_desc == "$16,000 through $24,999":
            income_category_num=4
        elif income_desc == "$25,000 through $34,999":
            income_category_num=5
        elif income_desc == "$35,000 through $49,999":
            income_category_num=6
        elif income_desc == "$50,000 through $74,999":
            income_category_num=7
        elif income_desc == "$75,000 through $99,999":
            income_category_num=8
        elif income_desc == "$100,000 through $199,999":
            income_category_num=9
        elif income_desc == "$200,000 and greater":
            income_category_num=10

        drug_used = True
        if 'None' in drug_during_pregnancy:
            drug_used = False

        avg_screen_time = (screen_week*5 + screen_weekend*2)/7

        model = load_pretrained_model("model.sav")
        data = np.array([float(subject_bmi), float(sleep_hours), income_category_num, float(avg_screen_time), drug_used])
        proba, res = inference_model(model, data)
        col1, col2 = st.columns(2)
        if res == 1:
            st.write("The subject is likely to be in the high-risk ADHD group")
            col1.metric("Prediction", "High Risk")
        elif res == 0 and proba >=0.9:
            st.write("The subject is not likely to be in the high-risk ADHD group")
            col1.metric("Prediction", "Low Risk")
        else:
            st.write("Information is not sufficient to make a decision.")
            col1.metric("Prediction", "Undecided")
        st.markdown('## ****What Your Prediction Results Mean****')
        st.markdown('### If the subject is predicted to be in a “high-risk” group')
        st.markdown('- The model predicted that the subject is highly likely to be in a high-risk group of ADHD.')
        st.markdown('- The subject should consider consult with a doctor for accurate diagnosis.')
        st.markdown('### If you are predicted to be in a “low-risk” group')
        st.markdown('- The model predicted that the subject is less likely to be in a high-risk group of ADHD.')
        st.markdown('- But the test does not rule out that the subject is not in a high-risk group of ADHD.')
        st.markdown('- Consider taking the test once in a quarter. Multiple “low-risk” prediction increases the confidence of prediction.')
        st.markdown('### If you are predicted to be in a “undecided” group')
        st.markdown('- The model does not confident with it’s prediction.')
        st.markdown('- You should consider taking another test.')