File size: 8,677 Bytes
04e24ee
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# -*- coding: utf-8 -*-
"""Nigerian Car Price Model.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1RtrEB_oX2Q9llgG2KysiBNuIg-EEtpdv
"""

import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings("ignore")
sns.set_style("darkgrid")
sns.set_palette('RdYlGn')

#model
from sklearn.preprocessing import LabelEncoder,StandardScaler,MinMaxScaler
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error, r2_score
from sklearn.ensemble import RandomForestRegressor
from xgboost import XGBRegressor
from sklearn.linear_model import LinearRegression

import gradio as gr
import joblib

df = pd.read_csv("/content/Nigerian_Car_Prices.csv")

df.head()

df.info()

"""### Data Cleaning"""

df = df.drop('Build', axis = 1)

df = df.dropna()

df.shape

df['Price'] = df['Price'].str.replace(',', '') 
df['Price'] = df['Price'].astype(float)  

df['Year of manufacture'] = df['Year of manufacture'].astype(int)

df.describe()

"""### EDA

### Feature Engineering
"""

#the brand new is just 5, it will be drop
# Dropping the 'Brand New' category
df = df[df['Condition'] != 'Brand New']

X = df.drop(['Unnamed: 0', 'Price'], axis = 1)
y = df.Price

make_counts = X['Make'].value_counts()


# Get the values to replace with 'Others'
make_others = make_counts[make_counts < 14].index.tolist()

# Replace values with 'Others'
X['Make'] = X['Make'].apply(lambda x: 'Others' if x in make_others else x)

X_train,X_test, y_train,y_test = train_test_split(X,y, test_size = 0.2, random_state=10)


# Initializing the encoders and scaler for each column
make_encoder = LabelEncoder()
fuel_encoder = LabelEncoder()
transmission_encoder = LabelEncoder()
condition_encoder = LabelEncoder()
scaler = MinMaxScaler()

# Encoding and scaling each column individually
X_train['Make'] = make_encoder.fit_transform(X_train['Make'])
X_test['Make'] = make_encoder.transform(X_test['Make'])

X_train['Fuel'] = fuel_encoder.fit_transform(X_train['Fuel'])
X_test['Fuel'] = fuel_encoder.transform(X_test['Fuel'])

X_train['Transmission'] = transmission_encoder.fit_transform(X_train['Transmission'])
X_test['Transmission'] = transmission_encoder.transform(X_test['Transmission'])

X_train['Condition'] = condition_encoder.fit_transform(X_train['Condition'])
X_test['Condition'] = condition_encoder.transform(X_test['Condition'])

X_train[['Year of manufacture', 'Mileage', 'Engine Size']] = scaler.fit_transform(X_train[['Year of manufacture', 'Mileage', 'Engine Size']])
X_test[['Year of manufacture', 'Mileage', 'Engine Size']] = scaler.transform(X_test[['Year of manufacture', 'Mileage', 'Engine Size']])

# Save the encoders and scaler
joblib.dump(make_encoder, "make_encoder.joblib",compress=3)
joblib.dump(fuel_encoder, "fuel_encoder.joblib",compress=3)
joblib.dump(transmission_encoder, "transmission_encoder.joblib",compress=3)
joblib.dump(condition_encoder, "condition_encoder.joblib",compress=3)
joblib.dump(scaler, "scaler.joblib",compress=3)

"""#### Needed Model"""

# Initialize the models
rf_model = RandomForestRegressor(random_state=42)
xgb_model = XGBRegressor(random_state=42)
lr_model = LinearRegression()

# Fit the models on the training data
rf_model.fit(X_train, y_train)
xgb_model.fit(X_train, y_train)
lr_model.fit(X_train, y_train)

# Make predictions on the testing data
rf_preds = rf_model.predict(X_test)
xgb_preds = xgb_model.predict(X_test)
lr_preds = lr_model.predict(X_test)

# Evaluate the models using root mean squared error (RMSE)
rf_rmse = mean_squared_error(y_test, rf_preds, squared=False)
xgb_rmse = mean_squared_error(y_test, xgb_preds, squared=False)
lr_rmse = mean_squared_error(y_test, lr_preds, squared=False)

# Print the RMSE scores
print(f"Random Forest RMSE: {rf_rmse:.2f}")
print(f"XGBoost RMSE: {xgb_rmse:.2f}")
print(f"Linear Regression RMSE: {lr_rmse:.2f}")

#  R2 score
rf_r2 = r2_score(y_test, rf_preds)
print("Random Forest R2 Score:", rf_r2)


xgb_r2 = r2_score(y_test, xgb_preds)
print("XGBoost R2 Score:", xgb_r2)


lr_r2 = r2_score(y_test, lr_preds)
print("Linear Regression R2 Score:", lr_r2)

joblib.dump(xgb_model, "car_model.joblib", compress=3)

"""**Note: Many Models have been built, but only the needed ones were kept**"""

sns.histplot(xgb_preds, label='prediction',color='red')
sns.histplot(y_test, label='actual price', color = 'blue')
plt.title('Prediction Vs Actual')
plt.legend()
plt.show()

"""### Prediction"""

import joblib
def predict_car_price(make, year, condition, mileage, engine_size, fuel, transmission):
    # Load the encoders and scaler
    make_encoder = joblib.load("make_encoder.joblib")
    fuel_encoder = joblib.load("fuel_encoder.joblib")
    transmission_encoder = joblib.load("transmission_encoder.joblib")
    condition_encoder = joblib.load("condition_encoder.joblib")
    scaler = joblib.load("scaler.joblib")

    # Preprocess the input
    make_encoded = make_encoder.transform([make])[0]
    numerical_value = scaler.transform([[year,mileage, engine_size]])
    year_scaled = numerical_value[0][0]
    mileage_scaled = numerical_value[0][1]
    engine_size_scaled = numerical_value[0][2]
    fuel_encoded = fuel_encoder.transform([fuel])[0]
    condition_encoded = condition_encoder.transform([condition])[0]
    transmission_encoded = transmission_encoder.transform([transmission])[0]

    input_data = [[make_encoded, year_scaled, condition_encoded, mileage_scaled, engine_size_scaled, fuel_encoded, transmission_encoded]]
    input_df = pd.DataFrame(input_data, columns=['Make', 'Year of manufacture', 'Condition', 'Mileage', 'Engine Size', 'Fuel', 'Transmission'])

    # Make predictions
    predicted_price = xgb_model.predict(input_df)
    return round(predicted_price[0], 2)

predict_car_price('Toyota', 2010,'Nigerian Used', 3000, 2300, 'Petrol', 'Automatic')

"""### Gradio Interface"""

import gradio as gr
import joblib
def predict_car_price(make, year, condition, mileage, engine_size, fuel, transmission):
    # Load the encoders and scaler
    make_encoder = joblib.load("make_encoder.joblib")
    fuel_encoder = joblib.load("fuel_encoder.joblib")
    transmission_encoder = joblib.load("transmission_encoder.joblib")
    condition_encoder = joblib.load("condition_encoder.joblib")
    scaler = joblib.load("scaler.joblib")

    make_encoded = make_encoder.transform([make])[0]
    numerical_value = scaler.transform([[year,mileage, engine_size]])
    year_scaled = numerical_value[0][0]
    mileage_scaled = numerical_value[0][1]
    engine_size_scaled = numerical_value[0][2]
    fuel_encoded = fuel_encoder.transform([fuel])[0]
    condition_encoded = condition_encoder.transform([condition])[0]
    transmission_encoded = transmission_encoder.transform([transmission])[0]
    input_data = [[make_encoded, year_scaled, condition_encoded, mileage_scaled, engine_size_scaled, fuel_encoded, transmission_encoded]]
    input_df = pd.DataFrame(input_data, columns=['Make', 'Year of manufacture', 'Condition', 'Mileage', 'Engine Size', 'Fuel', 'Transmission'])

    # Make predictions
    predicted_price = xgb_model.predict(input_df)
    return round(predicted_price[0], 2)
make_dropdown = gr.inputs.Dropdown(['Acura', 'Audi', 'BMW', 'Chevrolet', 'Dodge', 'Ford', 'Honda',
       'Hyundai', 'Infiniti', 'Kia', 'Land Rover', 'Lexus', 'Mazda',
       'Mercedes-Benz', 'Mitsubishi', 'Nissan', 'Peugeot',
       'Pontiac', 'Toyota', 'Volkswagen', 'Volvo'], label="Make")
condition_dropdown = gr.inputs.Dropdown(['Foreign Used', 'Nigerian Used'], label="Condition")
fuel_dropdown = gr.inputs.Dropdown(["Petrol", "Diesel", "Electric"], label="Fuel")
transmission_dropdown = gr.inputs.Dropdown(["Manual", "Automatic", "AMT"], label="Transmission")
year_slider = gr.inputs.Slider(minimum=1992, maximum=2021, step=1, default=2010, label="Year")
mileage_slider = gr.inputs.Slider(minimum=1, maximum=300000, step=10, default=80000, label="Mileage")
engine_size_slider = gr.inputs.Slider(minimum=1, maximum=20000, step=1, default=100, label="Engine Size")

iface = gr.Interface(
fn=predict_car_price,
inputs=[make_dropdown, year_slider, condition_dropdown, mileage_slider, engine_size_slider, fuel_dropdown, transmission_dropdown],
outputs="number",
title="Car Price Prediction",
    description="Predict the price of a car based on its details, in Naira.",
    examples=[
        ["Toyota", 2010, "Nigerian Used", 80000, 2.0, "Petrol", "Automatic"],
        ["Mercedes-Benz", 2015, "Foreign Used", 50000, 1000, "Diesel", "AMT"],
    ],css=".gradio-container {background-color: lightgreen}"
)

iface.launch(share = True)