|
import pickle |
|
import pandas as pd |
|
import shap |
|
from shap.plots._force_matplotlib import draw_additive_plot |
|
import gradio as gr |
|
import numpy as np |
|
import matplotlib.pyplot as plt |
|
|
|
|
|
loaded_model = pickle.load(open("heart_ba4522_example.pkl", 'rb')) |
|
|
|
|
|
explainer = shap.Explainer(loaded_model) |
|
|
|
|
|
def main_func(age, sex, cp, trestbps, chol, fbs, restecg, thalach, |
|
exang, oldpeak, slope, ca, thal): |
|
new_row = pd.DataFrame.from_dict({'age': age, |
|
'sex':sex, |
|
'cp':cp, |
|
'trestbps':trestbps, |
|
'chol':chol, |
|
'fbs':fbs, |
|
'restecg':restecg, |
|
'thalach':thalach, |
|
'exang':exang, |
|
'oldpeak':oldpeak, |
|
'slope':slope, |
|
'ca':ca, |
|
'thal':thal |
|
}, orient = 'index').transpose() |
|
|
|
prob = loaded_model.predict_proba(new_row) |
|
|
|
shap_values = explainer(new_row) |
|
|
|
|
|
plot = shap.plots.bar(shap_values[0], max_display=7, order=shap.Explanation.abs, show_data='auto', show=False) |
|
|
|
plt.tight_layout() |
|
local_plot = plt.gcf() |
|
plt.rcParams['figure.figsize'] = 7,4 |
|
plt.close() |
|
|
|
return {"Normal Heart Condition": float(prob[0][0]), "Critical Heart Condition": 1-float(prob[0][0])}, local_plot |
|
|
|
|
|
title = "**Heart Condition Predictor & Interpreter** 🪐" |
|
description1 = """ |
|
This app takes inputs about patients' demographics and medical history to predict whether the patient has heart condition. There are two outputs from the app: 1- the predicted probability of normal condition or heart condition, 2- Shapley's force-plot which visualizes the extent to which each factor impacts the prediction. |
|
""" |
|
|
|
description2 = """ |
|
To use the app, click on one of the examples, or adjust the values of the patient factors, and click on Analyze. ✨ |
|
""" |
|
|
|
with gr.Blocks(title=title) as demo: |
|
gr.Markdown(f"## {title}") |
|
|
|
gr.Markdown(description1) |
|
gr.Markdown("""---""") |
|
gr.Markdown(description2) |
|
gr.Markdown("""---""") |
|
with gr.Row(): |
|
with gr.Column(): |
|
age = gr.Slider(label="age", minimum=17, maximum=74, value=24, step=1) |
|
sex = gr.Slider(label="sex", minimum=0, maximum=1, value=1, step=1) |
|
cp = gr.Slider(label="cp Score", minimum=1, maximum=4, value=3, step=.1) |
|
trestbps = gr.Slider(label="trestbps Score", minimum=94, maximum=200, value=150, step=.1) |
|
chol = gr.Slider(label="chol Score", minimum=126, maximum=564, value=400, step=.1) |
|
fbs = gr.Slider(label="fbs Score", minimum=0, maximum=1, value=0, step=.1) |
|
restecg = gr.Slider(label="restecg Score", minimum=0, maximum=2, value=1, step=.1) |
|
thalach = gr.Slider(label="thalach Score", minimum=71, maximum=202, value=90, step=.1) |
|
exang = gr.Slider(label="exang Score", minimum=0, maximum=1, value=1, step=.1) |
|
oldpeak = gr.Slider(label="oldpeak Score", minimum=0, maximum=6, value=4, step=.1) |
|
slope = gr.Slider(label="slope Score", minimum=1, maximum=3, value=2, step=.1) |
|
ca = gr.Slider(label="ca Score", minimum=0, maximum=3, value=2, step=.1) |
|
thal = gr.Slider(label="thal Score", minimum=3, maximum=7, value=4, step=.1) |
|
|
|
submit_btn = gr.Button("Analyze") |
|
with gr.Column(visible=True,scale=1, min_width=600) as output_col: |
|
label = gr.Label(label = "Predicted Label") |
|
local_plot = gr.Plot(label = 'Shap:') |
|
|
|
submit_btn.click( |
|
main_func, |
|
[age, sex,cp,trestbps,chol,fbs,restecg,thalach,exang,oldpeak,slope,ca,thal], |
|
[label,local_plot], api_name="Heart_Condition" |
|
) |
|
|
|
gr.Markdown("### Click on any of the examples below to see how it works:") |
|
gr.Examples([[33,0,1,100,230,1,1,150,0,.9,2,1,6], [39,1,0,170,200,1,1,150,0,1.4,2,1,6]], |
|
[age,sex,cp,trestbps,chol,fbs,restecg,thalach,exang,oldpeak,slope,ca,thal], |
|
[label,local_plot], main_func, cache_examples=True) |
|
|
|
demo.launch() |
