ML_DEMO / app_function.py
cnasa's picture
Update app_function.py
2bb3427
Raw History Blame Contribute Delete
40.4 kB
import pickle import pandas as pd import sklearn import numpy as np import gradio as gr import imblearn import sklearn from sklearn.ensemble import RandomForestClassifier from sklearn.model_selection import train_test_split from sklearn.metrics import confusion_matrix,plot_confusion_matrix,classification_report from imblearn.over_sampling import RandomOverSampler from sklearn.preprocessing import StandardScaler,OneHotEncoder from sklearn.compose import make_column_transformer import tensorflow as tf from tensorflow.keras.applications import imagenet_utils from tensorflow.keras.utils import img_to_array from tensorflow.keras.models import load_model import pdfplumber import re from collections import namedtuple from compress_pickle import dump, load import warnings warnings.filterwarnings("ignore") # Load Data def loading_data(): df = pd.read_csv("diabetes_012__health_indicators_BRFSS2015.csv") #convert all columns to integer for col in df.columns: df[col] = df[col].astype("int") # Drop duplicated rows df_ = df.drop_duplicates() df=df_ X,y = df.drop(['Diabetes_012'],axis=1),df['Diabetes_012'].values classes = np.unique(y) #Oversampling data randomSampler = RandomOverSampler(sampling_strategy='all',random_state=24) X_new,y_new = randomSampler.fit_resample(X,y) new_df = X_new.copy() new_df["Diabetes_012"] = y_new Xtrain,Xtest,Ytrain,Ytest = train_test_split(X_new,y_new,test_size=0.2,random_state=24,stratify=y_new) # Some feature engineering Xtrain_transf = Xtrain.copy() Xtrain_transf["Age2"] = Xtrain_transf["Age"]**2 ## Numerical column transformation col_num=["BMI","MentHlth","PhysHlth","Age","Age2"] num_col_trans = make_column_transformer((StandardScaler(),col_num),remainder="passthrough") Xtrain_transf_std = num_col_trans.fit_transform(Xtrain_transf,Ytrain) Xtrain_transf_std = pd.DataFrame(Xtrain_transf_std ,columns=list(Xtrain_transf.columns) ) # Données tests: Xtest_transf = Xtest.copy() Xtest_transf["Age2"] = Xtest_transf["Age"]**2 Xtest_transf_std = num_col_trans.transform(Xtest_transf) Xtest_transf_std = pd.DataFrame(Xtest_transf_std ,columns=list(Xtest_transf.columns) ) ## Categorical columns transformation col_cat = ["GenHlth","Education","Income"] cat_col_trans = make_column_transformer((OneHotEncoder(handle_unknown = 'ignore'),col_cat),remainder="passthrough") Xtrain_transf_std_encoded = cat_col_trans.fit_transform(Xtrain_transf_std,Ytrain) # Sur les données tests : Xtest_encoded = cat_col_trans.transform(Xtest_transf_std) results = {"num_col_trans": num_col_trans, "cat_col_trans": cat_col_trans} return results def load_models(): ## Load Model from pickle file with open("Model_Package", "rb") as f: model = load(f,compression='gzip') res = {"model": model} return res # Function use to get the user input data def user_input(HighBP,HighChol, CholCheck, BMI, Smoker,Stroke,HeartDiseaseorAttack, PhysActivity, Fruits, Veggies,HvyAlcoholConsump,AnyHealthcare, NoDocbcCost, GenHlth, MentHlth, PhysHlth, DiffWalk, Sex, Age, Education,Income): # Preprare data type AnyHealthcare = 0 if AnyHealthcare=="Non" else 1 HvyAlcoholConsump = 0 if HvyAlcoholConsump=="Non" else 1 HighBP = 0 if HighBP=="Non" else 1 NoDocbcCost = 0 if NoDocbcCost=="Non" else 1 HighChol= 0 if HighChol=="Non" else 1 GenHlth = int(GenHlth) CholCheck = 0 if CholCheck=="Non" else 1 MentHlth = int(MentHlth) BMI = int(BMI) PhysHlth = int(PhysHlth) Smoker = 0 if Smoker=="Non" else 1 DiffWalk = 0 if DiffWalk=="Non" else 1 Stroke = 0 if Stroke=="Non" else 1 Sex = 0 if Sex=="Homme" else 1 HeartDiseaseorAttack = 0 if HeartDiseaseorAttack=="Non" else 1 Age = int(Age) PhysActivity = 0 if PhysActivity=="Non" else 1 Education = int(Education) Fruits = 0 if Fruits=="Non" else 1 Income = int(Income) Veggies = 0 if Veggies=="Non" else 1 user_data = {'HighBP':HighBP, 'HighChol':HighChol, 'CholCheck':CholCheck, 'BMI':BMI, 'Smoker':Smoker, 'Stroke':Stroke, 'HeartDiseaseorAttack':HeartDiseaseorAttack, 'PhysActivity':PhysActivity, 'Fruits':Fruits, 'Veggies':Veggies, 'HvyAlcoholConsump':HvyAlcoholConsump, 'AnyHealthcare':AnyHealthcare, 'NoDocbcCost':NoDocbcCost, 'GenHlth':GenHlth, 'MentHlth':MentHlth, 'PhysHlth':PhysHlth, 'DiffWalk':DiffWalk, 'Sex':Sex, 'Age':Age, 'Education':Education, 'Income':Income} user_df = pd.DataFrame(user_data,index=[0]) # ## Load Model from pickle file MODELS = load_models() model = MODELS["model"] transformers = loading_data() num_col_trans = transformers["num_col_trans"] cat_col_trans = transformers["cat_col_trans"] # Data preparation before fitting to the model user_df["Age2"] = user_df["Age"]**2 user_df_std = num_col_trans.transform(user_df) user_df_std = pd.DataFrame(user_df_std ,columns=list(user_df.columns) ) user_df_encoded = cat_col_trans.transform(user_df_std) user_encoded = np.array(user_df_encoded).reshape((1,-1)) #PREDICTION target_names = ['No diabetes', 'Prediabetes', 'Diabetes'] pred_proba = model.predict_proba(user_encoded)[0] results = {classe : pred_proba[i] for i, classe in enumerate(target_names)} max_prob = 0 ma_classe = "" for k,v in results.items(): if v > max_prob: max_prob= v ma_classe= k STYLE = """ <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-0evHe/X+R7YkIZDRvuzKMRqM+OrBnVFBL6DOitfPri4tjfHxaWutUpFmBp4vmVor" crossorigin="anonymous"> """ OUTPUT_OK = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-info mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-success">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-success">Félicitation ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) OUTPUT_PREDANGER = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-info mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <h5 class="card-title"> </h5> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-warning">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-warning">Warning ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) OUTPUT_DANGER = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-info mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <h5 class="card-title"> </h5> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-danger">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-danger">Danger ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) if ma_classe=="No diabetes" and max_prob > 0.7 : max_prob = OUTPUT_OK elif (ma_classe=="No diabetes" and max_prob <= 0.7) or ma_classe=="Prediabetes" : max_prob = OUTPUT_PREDANGER else: max_prob = OUTPUT_DANGER return max_prob # Get PDF File for user def get_data_inputPDF(PDF_file): path_file = PDF_file.name with pdfplumber.open(path_file) as pdf: num_page = 0 page = pdf.pages[num_page] text = page.extract_text(x_tolerance=2, y_tolerance=0) AnyHealthcare_pattern = re.compile('^(AnyHealthcare)\s+(\w{3})') HighBP_pattern = re.compile('^(HighBP)\s+(\w{3})') NoDocbcCost_pattern = re.compile('^(NoDocbcCost)\s+(\w{3})') HighChol_pattern = re.compile('^(HighChol)\s+(\w{3})') GenHlth_pattern = re.compile('^(GenHlth)\s+(\d+)') CholCheck_pattern = re.compile('^(CholCheck)\s+(\w{3})') MentHlth_pattern = re.compile('^(MentHlth)\s+(\d+)') BMI_pattern = re.compile('^(BMI)\s+(\d+)') PhysHlth_pattern = re.compile('^(PhysHlth)\s+(\d+)') Smoker_pattern = re.compile('^(Smoker)\s+(\w{3})') DiffWalk_pattern = re.compile('^(DiffWalk)\s+(\w{3})') Stroke_pattern = re.compile('^(Stroke)\s+(\w{3})') Sex_pattern = re.compile('^(Sex)\s+(\w+)') HeartDiseaseorAttack_pattern = re.compile('^(HeartDiseaseorAttack)\s+(\w{3})') Age_pattern = re.compile('^(Age)\s+(\d+)') PhysActivity_pattern = re.compile('^(PhysActivity)\s+(\w{3})') Education_pattern = re.compile('^(Education)\s+(\d+)') Fruits_pattern = re.compile('^(Fruits)\s+(\w{3})') Income_pattern = re.compile('^(Income)\s+(\d+)') Veggies_pattern = re.compile('^(Veggies)\s+(\w{3})') HvyAlcoholConsump_pattern = re.compile('^(HvyAlcoholConsump)\s+(\w{3})') for line in text.split("\n"): AnyHealthcare_ligne = AnyHealthcare_pattern.search(line) HighBP_ligne = HighBP_pattern.search(line) NoDocbcCost_ligne = NoDocbcCost_pattern.search(line) HighChol_ligne = HighChol_pattern.search(line) GenHlth_ligne = GenHlth_pattern.search(line) CholCheck_ligne = CholCheck_pattern.search(line) MentHlth_ligne = MentHlth_pattern.search(line) BMI_ligne = BMI_pattern.search(line) PhysHlth_ligne = PhysHlth_pattern.search(line) Smoker_ligne = Smoker_pattern.search(line) DiffWalk_ligne = DiffWalk_pattern.search(line) Stroke_ligne = Stroke_pattern.search(line) Sex_ligne = Sex_pattern.search(line) HeartDiseaseorAttack_ligne = HeartDiseaseorAttack_pattern.search(line) Age_ligne = Age_pattern.search(line) PhysActivity_ligne = PhysActivity_pattern.search(line) Education_ligne = Education_pattern.search(line) Fruits_ligne = Fruits_pattern.search(line) Income_ligne = Income_pattern.search(line) Veggies_ligne = Veggies_pattern.search(line) HvyAlcoholConsump_ligne = HvyAlcoholConsump_pattern.search(line) if AnyHealthcare_ligne: AnyHealthcare = AnyHealthcare_ligne.group(2) if HvyAlcoholConsump_ligne: HvyAlcoholConsump = HvyAlcoholConsump_ligne.group(2) if HighBP_ligne: HighBP = HighBP_ligne.group(2) if NoDocbcCost_ligne: NoDocbcCost = NoDocbcCost_ligne.group(2) if HighChol_ligne: HighChol= HighChol_ligne.group(2) if GenHlth_ligne: GenHlth = GenHlth_ligne.group(2) if CholCheck_ligne: CholCheck = CholCheck_ligne.group(2) if MentHlth_ligne: MentHlth = MentHlth_ligne.group(2) if BMI_ligne: BMI = BMI_ligne.group(2) if PhysHlth_ligne: PhysHlth = PhysHlth_ligne.group(2) if Smoker_ligne: Smoker = Smoker_ligne.group(2) if DiffWalk_ligne: DiffWalk = DiffWalk_ligne.group(2) if Stroke_ligne: Stroke = Stroke_ligne.group(2) if Sex_ligne: Sex = Sex_ligne.group(2) if HeartDiseaseorAttack_ligne: HeartDiseaseorAttack = HeartDiseaseorAttack_ligne.group(2) if Age_ligne: Age = Age_ligne.group(2) if PhysActivity_ligne: PhysActivity = PhysActivity_ligne.group(2) if Education_ligne: Education = Education_ligne.group(2) if Fruits_ligne: Fruits = Fruits_ligne.group(2) if Income_ligne: Income = Income_ligne.group(2) if Veggies_ligne: Veggies = Veggies_ligne.group(2) tab = namedtuple('table','HighBP HighChol CholCheck BMI Smoker Stroke HeartDiseaseorAttack \ PhysActivity Fruits Veggies HvyAlcoholConsump AnyHealthcare NoDocbcCost GenHlth \ MentHlth PhysHlth DiffWalk \ Sex Age Education Income') data_unpreprared = pd.DataFrame([tab(HighBP,HighChol, CholCheck, BMI, Smoker,Stroke,HeartDiseaseorAttack, PhysActivity, Fruits, Veggies,HvyAlcoholConsump,AnyHealthcare, NoDocbcCost, GenHlth, MentHlth, PhysHlth, DiffWalk, Sex, Age, Education,Income)]) # Preprare data type AnyHealthcare = 0 if AnyHealthcare=="Non" else 1 HvyAlcoholConsump = 0 if HvyAlcoholConsump=="Non" else 1 HighBP = 0 if HighBP=="Non" else 1 NoDocbcCost = 0 if NoDocbcCost=="Non" else 1 HighChol= 0 if HighChol=="Non" else 1 GenHlth = int(GenHlth) CholCheck = 0 if CholCheck=="Non" else 1 MentHlth = int(MentHlth) BMI = int(BMI) PhysHlth = int(PhysHlth) Smoker = 0 if Smoker=="Non" else 1 DiffWalk = 0 if DiffWalk=="Non" else 1 Stroke = 0 if Stroke=="Non" else 1 Sex = 0 if Sex=="Homme" else 1 HeartDiseaseorAttack = 0 if HeartDiseaseorAttack=="Non" else 1 Age = int(Age) PhysActivity = 0 if PhysActivity=="Non" else 1 Education = int(Education) Fruits = 0 if Fruits=="Non" else 1 Income = int(Income) Veggies = 0 if Veggies=="Non" else 1 user_df = pd.DataFrame([tab(HighBP,HighChol, CholCheck, BMI, Smoker,Stroke,HeartDiseaseorAttack, PhysActivity, Fruits, Veggies,HvyAlcoholConsump,AnyHealthcare, NoDocbcCost, GenHlth, MentHlth, PhysHlth, DiffWalk, Sex, Age, Education,Income)]) # #LOAD MODEL AND TRANSFORMERS MODELS = load_models() model = MODELS["model"] transformers = loading_data() num_col_trans = transformers["num_col_trans"] cat_col_trans = transformers["cat_col_trans"] # Data preparation before fitting to the model user_df["Age2"] = user_df["Age"]**2 user_df_std = num_col_trans.transform(user_df) user_df_std = pd.DataFrame(user_df_std ,columns=list(user_df.columns) ) user_df_encoded = cat_col_trans.transform(user_df_std) user_encoded = np.array(user_df_encoded).reshape((1,-1)) #PREDICTION target_names = ['No diabetes', 'Prediabetes', 'Diabetes'] pred_proba = model.predict_proba(user_encoded)[0] results = {classe : pred_proba[i] for i, classe in enumerate(target_names)} max_prob = 0 ma_classe = "" for k,v in results.items(): if v > max_prob: max_prob= v ma_classe= k STYLE = """ <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-0evHe/X+R7YkIZDRvuzKMRqM+OrBnVFBL6DOitfPri4tjfHxaWutUpFmBp4vmVor" crossorigin="anonymous"> """ OUTPUT_OK = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-success">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-success">Félicitation ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) OUTPUT_PREDANGER = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-warning">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-warning">Warning ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) OUTPUT_DANGER = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-3 g-4"> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> No diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["No diabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-warning mb-3 text-center btn-success"> <div class="card-header"> Prediabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Prediabetes"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Diabetes </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Diabetes"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-danger">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-danger">Danger ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> <br> <div class="row"><h1 style="text-align: center"> <h1> Importance des variables: </h1> <img src="https://huggingface.co/spaces/cnasa/ML_DEMO/resolve/main/feature_importance.PNG" alt="feature importance"/> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) if ma_classe=="No diabetes" and max_prob > 0.7 : max_prob = OUTPUT_OK elif (ma_classe=="No diabetes" and max_prob <= 0.7) or ma_classe=="Prediabetes" : max_prob = OUTPUT_PREDANGER else: max_prob = OUTPUT_DANGER return data_unpreprared,max_prob def Prediction_VGG16(image): #Prepare image IMG_SIZE = 224 image = img_to_array(image) image = image*1.0/255.0 img_prepared = image.reshape((-1,IMG_SIZE,IMG_SIZE,3)) #Load model vgg6 package path = "model_vgg16.h5" my_model =load_model(path) #Prediction classes = ["Brain Tumor","Healthy"] prediction = my_model.predict(img_prepared)[0] prediction = prediction.tolist() results = {k:v for k,v in zip(classes,prediction)} max_prob = 0 ma_classe = "" for k,v in results.items(): if v > max_prob: max_prob= v ma_classe= k STYLE = """ <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-0evHe/X+R7YkIZDRvuzKMRqM+OrBnVFBL6DOitfPri4tjfHxaWutUpFmBp4vmVor" crossorigin="anonymous"> """ OUTPUT_OK = ( STYLE + f"""<html> <body> <div class="row row-cols-1 row-cols-md-2 g-4"> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Brain Tumor </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Brain Tumor"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> Healthy </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Healthy"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-success">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-success">Félicitation ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> <script> </script> </html> """ ) OUTPUT_PREDANGER = ( STYLE + f"""<html> <body> </div> <div class="row row-cols-1 row-cols-md-2 g-4"> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Brain Tumor </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Brain Tumor"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> Healthy </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Healthy"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-warning">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-warning">Warning ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer !</button> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) OUTPUT_DANGER = ( STYLE + f"""<html> <body> </div> <div class="row row-cols-1 row-cols-md-2 g-4"> <div class="col"> <div class="card text-bg-danger mb-3 text-center btn-success"> <div class="card-header"> Brain Tumor </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Brain Tumor"],2)*100}%</p> </div> </div> </div> <div class="col"> <div class="card text-bg-success mb-3 text-center btn-success"> <div class="card-header"> Healthy </div> <div class="card-body"> <p class="card-text fw-bold">{round(results["Healthy"],2)*100}%</p> </div> </div> </div> </div> <div class="container"> <div class="row"><h1 style="text-align: center">Le Patient est détecté <b> {ma_classe} </b> avec une probabilité de </h1></div> <div class="row"><h1 class="display-1 text-success" style="text-align: center"><button type="button" class="btn btn-danger">{round(max_prob,2)*100}%</button></h1></div> <div class="row"><h1 style="text-align: center">d'être dans cette classe </h1></div> <div class="row"><h1 style="text-align: center"> <button type="button" class="btn btn-danger">Danger ! </button> <br> <button type="button" class="btn btn-info">Vous pouvez ressayer!</button> </div> <script src="https://cdn.jsdelivr.net/npm/@popperjs/core@2.11.5/dist/umd/popper.min.js" integrity="sha384-Xe+8cL9oJa6tN/veChSP7q+mnSPaj5Bcu9mPX5F5xIGE0DVittaqT5lorf0EI7Vk" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.2.0-beta1/dist/js/bootstrap.min.js" integrity="sha384-kjU+l4N0Yf4ZOJErLsIcvOU2qSb74wXpOhqTvwVx3OElZRweTnQ6d31fXEoRD1Jy" crossorigin="anonymous"></script> </body> </html> """ ) if ma_classe=="Healthy" and max_prob > 0.7 : max_prob = OUTPUT_OK elif ma_classe=="Healthy" and max_prob <= 0.7: max_prob = OUTPUT_PREDANGER else: max_prob = OUTPUT_DANGER return max_prob
import pickle
import pandas as pd
import sklearn
import numpy as np
import gradio as gr
import imblearn
import sklearn
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix,plot_confusion_matrix,classification_report
from imblearn.over_sampling import RandomOverSampler
from sklearn.preprocessing import StandardScaler,OneHotEncoder
from sklearn.compose import make_column_transformer
import tensorflow as tf
from tensorflow.keras.applications import imagenet_utils
from tensorflow.keras.utils import img_to_array
from tensorflow.keras.models import load_model
import pdfplumber
import re
from collections import namedtuple
from compress_pickle import dump, load
import warnings
warnings.filterwarnings("ignore")
# Load Data
def loading_data():
df = pd.read_csv("diabetes_012__health_indicators_BRFSS2015.csv")
#convert all columns to integer
for col in df.columns:
df[col] = df[col].astype("int")
# Drop duplicated rows
df_ = df.drop_duplicates()
df=df_
X,y = df.drop(['Diabetes_012'],axis=1),df['Diabetes_012'].values
classes = np.unique(y)
#Oversampling data
randomSampler = RandomOverSampler(sampling_strategy='all',random_state=24)
X_new,y_new = randomSampler.fit_resample(X,y)
new_df = X_new.copy()
new_df["Diabetes_012"] = y_new
Xtrain,Xtest,Ytrain,Ytest = train_test_split(X_new,y_new,test_size=0.2,random_state=24,stratify=y_new)
# Some feature engineering
Xtrain_transf = Xtrain.copy()
Xtrain_transf["Age2"] = Xtrain_transf["Age"]**2
## Numerical column transformation
col_num=["BMI","MentHlth","PhysHlth","Age","Age2"]
num_col_trans = make_column_transformer((StandardScaler(),col_num),remainder="passthrough")
Xtrain_transf_std = num_col_trans.fit_transform(Xtrain_transf,Ytrain)
Xtrain_transf_std = pd.DataFrame(Xtrain_transf_std ,columns=list(Xtrain_transf.columns) )
# Données tests:
Xtest_transf = Xtest.copy()
Xtest_transf["Age2"] = Xtest_transf["Age"]**2
Xtest_transf_std = num_col_trans.transform(Xtest_transf)
Xtest_transf_std = pd.DataFrame(Xtest_transf_std ,columns=list(Xtest_transf.columns) )
## Categorical columns transformation
col_cat = ["GenHlth","Education","Income"]
cat_col_trans = make_column_transformer((OneHotEncoder(handle_unknown = 'ignore'),col_cat),remainder="passthrough")
Xtrain_transf_std_encoded = cat_col_trans.fit_transform(Xtrain_transf_std,Ytrain)
# Sur les données tests :
Xtest_encoded = cat_col_trans.transform(Xtest_transf_std)
results = {"num_col_trans": num_col_trans,
"cat_col_trans": cat_col_trans}
return results
def load_models():
## Load Model from pickle file
with open("Model_Package", "rb") as f:
model = load(f,compression='gzip')
res = {"model": model}
return res
# Function use to get the user input data