File size: 10,614 Bytes
7fa3b30 d6a735f 7fa3b30 d6a735f 7fa3b30 d6a735f 7fa3b30 a6d54a9 7fa3b30 d6a735f 7fa3b30 d6a735f 7fa3b30 d6a735f 7fa3b30 d6a735f 7fa3b30 d6a735f 268d97f d6a735f 268d97f 7fa3b30 5b14f9b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 |
import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.ensemble import RandomForestClassifier, VotingClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.naive_bayes import GaussianNB
from sklearn.neural_network import MLPClassifier
from sklearn.ensemble import GradientBoostingClassifier
from xgboost import XGBClassifier
from lightgbm import LGBMClassifier
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
st.title('Diabetes Prediction Application')
st.write('''
Please fill in the attributes below, then hit the Predict button
to get your results.
''')
st.header('Input Attributes')
age = st.slider('Your Age (Years)', min_value=0.0, max_value=100.0, value=50.0, step=1.0)
st.write(''' ''')
gen = st.radio("Your Gender", ('Male', 'Female'))
st.write(''' ''')
# gender conversion
if gen == "Male":
gender = 1
else:
gender = 0
urea = st.slider('Urea', min_value=0.0, max_value=100.0, value=50.0, step=0.1)
st.write(''' ''')
cr = st.slider('Creatinine Ratio(Cr)', min_value=0.0, max_value=1000.0, value=500.0, step=1.0)
st.write(''' ''')
hb = st.slider('HbA1c', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
st.write(''' ''')
chol = st.slider('Cholesterol (Chol)', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
st.write(''' ''')
tg = st.slider('Triglycerides(TG) Cholesterol', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
st.write(''' ''')
hdl = st.slider('HDL Cholesterol', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
st.write(''' ''')
ldl = st.slider('LDL Cholesterol', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
st.write(''' ''')
vldl = st.slider('VLDL Cholesterol', min_value=0.0, max_value=50.0, value=25.0, step=0.1)
st.write(''' ''')
bmi = st.slider('BMI', min_value=0.0, max_value=50.0, value=25.0, step=0.1)
st.write(''' ''')
selected_models = st.multiselect("Choose Classifier Models", ('Random Forest', 'Naïve Bayes', 'Logistic Regression', 'K-Nearest Neighbors', 'Decision Tree', 'Gradient Boosting', 'LightGBM', 'XGBoost', 'Multilayer Perceptron', 'Artificial Neural Network', 'Support Vector Machine'))
st.write(''' ''')
# Initialize an empty list to store the selected models
models_to_run = []
# Check which models were selected and add them to the models_to_run list
if 'Random Forest' in selected_models:
models_to_run.append(RandomForestClassifier())
if 'Naïve Bayes' in selected_models:
models_to_run.append(GaussianNB())
if 'Logistic Regression' in selected_models:
models_to_run.append(LogisticRegression())
if 'K-Nearest Neighbors' in selected_models:
models_to_run.append(KNeighborsClassifier())
if 'Decision Tree' in selected_models:
models_to_run.append(DecisionTreeClassifier())
if 'Gradient Boosting' in selected_models:
models_to_run.append(GradientBoostingClassifier())
if 'Support Vector Machine' in selected_models:
models_to_run.append(SVC(probability=True))
if 'LightGBM' in selected_models:
models_to_run.append(LGBMClassifier())
if 'XGBoost' in selected_models:
models_to_run.append(XGBClassifier())
if 'Multilayer Perceptron' in selected_models:
models_to_run.append(MLPClassifier())
if 'Artificial Neural Network' in selected_models:
models_to_run.append(MLPClassifier(hidden_layer_sizes=(100,), max_iter=100))
user_input = np.array([age, gender, urea, cr, hb, chol, tg, hdl, vldl,
ldl, bmi]).reshape(1, -1)
# import dataset
def get_dataset():
data = pd.read_csv('updated_diabetes.csv')
# Transforming class into numerical format
data['CLASS'] = data['CLASS'].apply(lambda x: 0 if x == 'N' else 1)
# Transforming Gender into numerical format
data['Gender'] = data['Gender'].apply(lambda x: 1 if x == 'M' else 0)
# Calculate the correlation matrix
# corr_matrix = data.corr()
# Create a heatmap of the correlation matrix
# plt.figure(figsize=(10, 8))
# sns.heatmap(corr_matrix, annot=True, cmap='coolwarm')
# plt.title('Correlation Matrix')
# plt.xticks(rotation=45)
# plt.yticks(rotation=0)
# plt.tight_layout()
# Display the heatmap in Streamlit
# st.pyplot()
return data
def generate_model_labels(model_names):
model_labels = []
for name in model_names:
words = name.split()
if len(words) > 1:
# Multiple words, use initials
label = "".join(word[0] for word in words)
else:
# Single word, take the first 3 letters
label = name[:3]
model_labels.append(label)
return model_labels
if st.button('Submit'):
df = get_dataset()
# fix column names
df.columns = (["id", "pation_no", "gender", "age", "urea", "cr",
"hb", "chol", "tg", "hdl", "ldl",
"vldl", "bmi", "target"])
# Split the dataset into train and test
X = df.drop(['target','id','pation_no'], axis=1)
y = df['target']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# Create two columns to divide the screen
left_column, right_column = st.columns(2)
# Left column content
with left_column:
# Create a VotingClassifier with the top 3 models
ensemble = VotingClassifier(
estimators=[('rf', RandomForestClassifier()), ('xgb', XGBClassifier()), ('gb', LGBMClassifier())],
voting='soft')
# Fit the voting classifier to the training data
ensemble.fit(X_train, y_train)
# Make predictions on the test set
model_predictions = ensemble.predict(user_input)
model_prob = ensemble.predict_proba(user_input)[:, 1]
# Evaluate the model's performance on the test set
ensamble_accuracy = accuracy_score(y_test, ensemble.predict(X_test))
ensamble_precision = precision_score(y_test, ensemble.predict(X_test))
ensamble_recall = recall_score(y_test, ensemble.predict(X_test))
ensamble_f1score = f1_score(y_test, ensemble.predict(X_test))
if model_predictions == 1:
st.write(f'According to Ensemble Model You have a **Very High Chance (1)** of Diabetes.')
st.write(f'Diabetes Probability: ', (model_prob* 100))
else:
st.write(f'According to Ensemble Model You have a **Very Low Chance (0)** of Diabetes.')
st.write(f'Diabetes Probability: ', (model_prob* 100))
st.write('Ensemble Model Accuracy:', ensamble_accuracy)
st.write('Ensemble Model Precision:', ensamble_precision)
st.write('Ensemble Model Recall:', ensamble_recall)
st.write('Ensemble Model F1 Score:', ensamble_f1score)
st.write('------------------------------------------------------------------------------------------------------')
# Right column content
with right_column:
for model in models_to_run:
# Train the selected model
model.fit(X_train, y_train)
# Make predictions on the test set
model_predictions = model.predict(user_input)
model_prob = model.predict_proba(user_input)[:, 1]
# Evaluate the model's performance on the test set
model_accuracy = accuracy_score(y_test, model.predict(X_test))
model_precision = precision_score(y_test, model.predict(X_test))
model_recall = recall_score(y_test, model.predict(X_test))
model_f1score = f1_score(y_test, model.predict(X_test))
if model_predictions == 1:
st.write(f'According to {type(model).__name__} Model You have a **Very High Chance (1)** of Diabetes.')
st.write(f'Diabetes Probability: ', (model_prob* 100))
else:
st.write(f'According to {type(model).__name__} Model You have a **Very Low Chance (0)** of Diabetes.')
st.write(f'Diabetes Probability: ', (model_prob* 100))
st.write(f'{type(model).__name__} Accuracy:', model_accuracy)
st.write(f'{type(model).__name__} Precision:', model_precision)
st.write(f'{type(model).__name__} Recall:', model_recall)
st.write(f'{type(model).__name__} F1 Score:', model_f1score)
st.write('------------------------------------------------------------------------------------------------------')
# Initialize lists to store model names and their respective performance metrics
model_names = ['Ensemble']
accuracies = [ensamble_accuracy]
precisions = [ensamble_precision]
recalls = [ensamble_recall]
f1_scores = [ensamble_f1score]
# Loop through the selected models to compute their performance metrics
for model in models_to_run:
model_names.append(type(model).__name__)
model.fit(X_train, y_train)
model_predictions = model.predict(X_test)
accuracies.append(accuracy_score(y_test, model_predictions))
precisions.append(precision_score(y_test, model_predictions))
recalls.append(recall_score(y_test, model_predictions))
f1_scores.append(f1_score(y_test, model_predictions))
# Create a DataFrame to store the performance metrics
metrics_df = pd.DataFrame({
'Model': model_names,
'Accuracy': accuracies,
'Precision': precisions,
'Recall': recalls,
'F1 Score': f1_scores
})
# Get the model labels
model_labels = generate_model_labels(metrics_df['Model'])
# Plot the comparison graphs
plt.figure(figsize=(12, 10))
# Accuracy comparison
plt.subplot(2, 2, 1)
plt.bar(model_labels, metrics_df['Accuracy'], color='skyblue')
plt.title('Accuracy Comparison')
plt.ylim(0, 1)
# Precision comparison
plt.subplot(2, 2, 2)
plt.bar(model_labels, metrics_df['Precision'], color='orange')
plt.title('Precision Comparison')
plt.ylim(0, 1)
# Recall comparison
plt.subplot(2, 2, 3)
plt.bar(model_labels, metrics_df['Recall'], color='green')
plt.title('Recall Comparison')
plt.ylim(0, 1)
# F1 Score comparison
plt.subplot(2, 2, 4)
plt.bar(model_labels, metrics_df['F1 Score'], color='purple')
plt.title('F1 Score Comparison')
plt.ylim(0, 1)
# Adjust layout to prevent overlapping of titles
plt.tight_layout()
# Display the graphs in Streamlit
st.pyplot()
|