Spaces:
Sleeping
Sleeping
File size: 10,893 Bytes
03e39f5 | 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 | import gradio as gr
import numpy as np
import pickle
import os
import sqlite3
def create_database():
# Connecting to the SQLite database
conn = sqlite3.connect('churndatabase.db')
cursor = conn.cursor()
# Create customer_churn table
cursor.execute("""
CREATE TABLE customer_churn (
customer_id INT PRIMARY KEY,
gender VARCHAR(10),
age INT,
marital_status VARCHAR(10),
dependents INT,
contract_type VARCHAR(10),
internet_service VARCHAR(20),
phone_service VARCHAR(3),
multiple_lines VARCHAR(3),
online_security VARCHAR(3),
online_backup VARCHAR(3),
device_protection VARCHAR(3),
tech_support VARCHAR(3),
streaming_tv VARCHAR(3),
streaming_movies VARCHAR(3),
monthly_charges NUMBER(8, 2),
total_charges NUMBER(10, 2),
churn_status VARCHAR(3)
)
""")
# Create customer_churn_full table
cursor.execute("""
CREATE TABLE customer_churn_full (
customer_id INT PRIMARY KEY,
gender VARCHAR(10),
age INT,
marital_status VARCHAR(10),
dependents INT,
contract_type VARCHAR(10),
internet_service VARCHAR(20),
phone_service VARCHAR(3),
multiple_lines VARCHAR(3),
online_security VARCHAR(3),
online_backup VARCHAR(3),
device_protection VARCHAR(3),
tech_support VARCHAR(3),
streaming_tv VARCHAR(3),
streaming_movies VARCHAR(3),
monthly_charges NUMBER(8, 2),
total_charges NUMBER(10, 2),
churn_status VARCHAR(3),
call_duration_minutes INT,
latitude NUMBER(9, 6),
longitude NUMBER(9, 6)
)
""")
# Insert data into customer_churn table
cursor.execute("""
INSERT INTO customer_churn
SELECT
TRUNC(RANDOM() * 1000000) AS customer_id,
CASE WHEN RANDOM() < 0.5 THEN 'Male' ELSE 'Female' END AS gender,
TRUNC(RANDOM() * 60 + 18) AS age,
CASE WHEN RANDOM() < 0.5 THEN 'Married' ELSE 'Single' END AS marital_status,
TRUNC(RANDOM() * 5) AS dependents,
CASE WHEN RANDOM() < 0.5 THEN 'Monthly' ELSE 'Yearly' END AS contract_type,
CASE WHEN RANDOM() < 0.5 THEN 'DSL' ELSE 'Fiber Optic' END AS internet_service,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS phone_service,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS multiple_lines,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS online_security,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS online_backup,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS device_protection,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS tech_support,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS streaming_tv,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS streaming_movies,
ROUND(RANDOM() * 100, 2) AS monthly_charges,
ROUND(RANDOM() * 1000, 2) AS total_charges,
CASE WHEN RANDOM() < 0.2 THEN 'Yes' ELSE 'No' END AS churn_status
FROM
(SELECT 1)
CROSS JOIN
(SELECT 1)
LIMIT
500
""")
# Insert data into customer_churn_full table
cursor.execute("""
INSERT INTO customer_churn_full
SELECT
TRUNC(RANDOM() * 1000000) AS customer_id,
CASE WHEN RANDOM() < 0.5 THEN 'Male' ELSE 'Female' END AS gender,
TRUNC(RANDOM() * 60 + 18) AS age,
CASE WHEN RANDOM() < 0.5 THEN 'Married' ELSE 'Single' END AS marital_status,
TRUNC(RANDOM() * 5) AS dependents,
CASE WHEN RANDOM() < 0.5 THEN 'Monthly' ELSE 'Yearly' END AS contract_type,
CASE WHEN RANDOM() < 0.5 THEN 'DSL' ELSE 'Fiber Optic' END AS internet_service,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS phone_service,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS multiple_lines,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS online_security,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS online_backup,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS device_protection,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS tech_support,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS streaming_tv,
CASE WHEN RANDOM() < 0.5 THEN 'Yes' ELSE 'No' END AS streaming_movies,
ROUND(RANDOM() * 100, 2) AS monthly_charges,
ROUND(RANDOM() * 1000, 2) AS total_charges,
CASE WHEN RANDOM() < 0.2 THEN 'Yes' ELSE 'No' END AS churn_status,
TRUNC(RANDOM() * 1200) AS call_duration_minutes,
ROUND(RANDOM() * 180 - 90, 6) AS latitude,
ROUND(RANDOM() * 360 - 180, 6) AS longitude
FROM
(SELECT 1)
CROSS JOIN
(SELECT 1)
LIMIT
500
""")
conn.commit()
conn.close()
create_database()
#Loading ML Components
DIRPATH = os.path.dirname(os.path.realpath(__file__))
ml_join= os.path.join(DIRPATH,'ml_components.pkl')
with open(ml_join,"rb") as f:
loaded_model_components = pickle.load(f)
# Extracting ML components
model = loaded_model_components['model']
encoder = loaded_model_components['encoder']
scaler = loaded_model_components['scaler']
#To predict churn
def predict_churn(gender, SeniorCitizen, Partner, Dependents, PhoneService, MultipleLines,
InternetService, OnlineSecurity, OnlineBackup, DeviceProtection, TechSupport,
StreamingTV, StreamingMovies, Contract, PaperlessBilling, PaymentMethod,
tenure, MonthlyCharges, TotalCharges):
#Encoding categorical features
encoded_data = encoder.transform([[gender, SeniorCitizen, Partner, Dependents, PhoneService, MultipleLines,
InternetService, OnlineSecurity, OnlineBackup, DeviceProtection, TechSupport,
StreamingTV, StreamingMovies, Contract, PaperlessBilling, PaymentMethod]])
num_data = np.array([[tenure, MonthlyCharges, TotalCharges]])
scaled_num_data = scaler.transform(num_data)
combined_data = np.hstack((encoded_data, scaled_num_data))
#Making predictions using fitted model
model_output = model.predict_proba(combined_data)
prob_churn = float(model_output[0][1])
return {'Churn Probability': prob_churn}
with gr.Blocks(css="""
background-color: #6db05d;
padding: 2rem;
border: 2px solid #e86347;
}
.gradio-input {
border: 1px solid #ccc000;
padding: 0.5rem;
border-radius: 0.5rem;
font-size: 1rem;
}
.gradio-label {
font-weight: bold;
margin-bottom: 0.5rem;
}
.gradio-markdown h1 {
font-size: 4rem;
font-family: 'Georgia', serif;
margin-top: 0;
margin-bottom: 0.5rem;
color: white;
background-color: #1E90FF;
padding: 1rem;
border-radius: 0.5rem;
text-align: center;
}
.gradio-markdown h2 {
font-size: 2rem;
font-family: 'DejaVu Sans Mono', monospace;
text-align: center;
margin-top: 0;
margin-bottom: 0.5rem;
color: white;
background-color: #4169E1;
padding: 0.5rem 1rem;
border-radius: 0.5rem;
}
.gradio-accordion {
background-color: #6db05d;
border-radius: 0.75rem;
padding: 1rem;
margin-bottom: 1rem;
text-align: left;
}
""") as test:
gr.Markdown("# CHURN PREDICTION")
gr.Markdown("## Using machine learning and Oracle SQL")
gr.Markdown("### By Hitesh Beeraka and Yash Singhvi")
with gr.Row():
with gr.Column(scale=2):
with gr.Accordion(label="Personal Information", open=False):
gender = gr.Radio(choices=['Male', 'Female'], label='Gender', interactive=True)
SeniorCitizen = gr.Radio(choices=['Yes', 'No'], label='SeniorCitizen', interactive=True)
Partner = gr.Radio(choices=['Yes', 'No'], label='Partner', interactive=True)
Dependents = gr.Radio(choices=['Yes', 'No'], label='Dependents', interactive=True)
with gr.Column(scale=2):
with gr.Accordion(label="Phone and Internet Services", open=False):
PhoneService = gr.Radio(choices=['Yes', 'No'], label='PhoneService', interactive=True)
MultipleLines = gr.Radio(choices=['Yes', 'No'], label='MultipleLines', interactive=True)
InternetService = gr.Radio(choices=['Fiber optic', 'No', 'DSL'], label='InternetService', interactive=True)
with gr.Row():
with gr.Column(scale=2):
with gr.Accordion(label="Security and Support Services", open=False):
OnlineSecurity = gr.Radio(choices=['Yes', 'No'], label='OnlineSecurity', interactive=True)
OnlineBackup = gr.Radio(choices=['Yes', 'No'], label='OnlineBackup', interactive=True)
DeviceProtection = gr.Radio(choices=['Yes', 'No'], label='DeviceProtection', interactive=True)
TechSupport = gr.Radio(choices=['Yes', 'No'], label='TechSupport', interactive=True)
with gr.Column(scale=2):
with gr.Accordion(label="Entertainment Services", open=False):
StreamingTV = gr.Radio(choices=['Yes', 'No'], label='StreamingTV', interactive=True)
StreamingMovies = gr.Radio(choices=['Yes', 'No'], label='StreamingMovies', interactive=True)
with gr.Row():
with gr.Column(scale=2):
with gr.Accordion(label="Contract and Billing Information", open=False):
Contract = gr.Radio(choices=['Month-to-month', 'One year', 'Two year'], label='Contract', interactive=True)
PaperlessBilling = gr.Radio(choices=['Yes', 'No'], label='PaperlessBilling', interactive=True)
PaymentMethod = gr.Radio(choices=['Electronic check', 'Mailed check', 'Credit card (automatic)', 'Bank transfer (automatic)'], label='PaymentMethod', interactive=True)
with gr.Column(scale=2):
with gr.Accordion(label="Charges and Tenure", open=False):
tenure = gr.Number(label='Tenure', interactive=True)
MonthlyCharges = gr.Number(label='MonthlyCharges', interactive=True)
TotalCharges = gr.Number(label='TotalCharges', interactive=True)
output = gr.Label(label="Churn Probability")
submit_btn = gr.Button("Predict")
submit_btn.click(
fn=predict_churn,
inputs=[gender, SeniorCitizen, Partner, Dependents, PhoneService, MultipleLines, InternetService, OnlineSecurity, OnlineBackup, DeviceProtection, TechSupport, StreamingTV, StreamingMovies, Contract, PaperlessBilling, PaymentMethod, tenure, MonthlyCharges, TotalCharges],
outputs=output
)
test.launch(inbrowser=True, show_error=True,share=True) |