carprice – mohammedyousefshaik.com

Project Details

Step 1 - If Python Not Installed

Open command prompt by typing cmd and give command python – – version, check

if you dont have python you may see like this

Goto https://www.python.org/downloads/ and download the latest python version for windows

After the installer is downloaded, double click and run, Install now and select check box to add path as shown

Now close the command prompt and open and type python – – version, you will be able to see version number

Step 2 - Check Environment Variable

Goto Search Type This PC and Right Click on Properties

Click on Advanced System Settings

Click on Environment Variables

Select Path and Edit, Ensure Python as shown is included as per your installation location

Step 3 - Install Python Libraries

pip install numpy pandas scikit-learn streamlit

Once python libraries are installed you will see like this

Step 4 - Install Jupyter Notebook

Check if Jupyter Notebook is installed

If not installed, run command pip install notebook

Now Verify if Jupyter is Installed, if installed you will see like this

Jupyter Notebook will open in browser or

Goto File > New > Notebook – Select Python Version and Select Check Box

Rename the file

Step 5 - Lets Start Working On Project

import the libraries and run

import pandas as pd
import numpy as np

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

Download The Dataset and

https://drive.google.com/file/d/1LhaWOs3V3cvFX-isSqL4-E9M9VJXb0Ua/view?usp=sharing

Download The Dataset and

Try reading the file by running this command

cars_data = pd.read_csv(‘Cardetails.csv’)

You may get this error, this is due to csv file is not placed in right location

Check the correct path

give command

import os

print(os.getcwd())

now at this location that comes, place your Cardetails.csv file

Rerun

give command

import os

print(os.getcwd())

now at this location , place your Cardetails.csv file and rerun command cars_data = pd.read_csv(‘Cardetails.csv’) ,it will run successfully

Lets kook at the data

give command cars_data.head()

to fetch top 5 rows

Lets Clean Data , remove unnecessary columns

cars_data.drop(columns=[‘torque’],inplace=True)

Dropping the column torque

Checking the DataSet Size Number of Rows and Columns

run command cars_data.shape

So we having 8128 rows and 12 columns

Step 6 - Starting Pre Processing of Data

Check for null records

run command cars_data.isnull().sum()

Remove the null records

run command cars_data.dropna(inplace=True)

Check for duplicate records

run command cars_data.isnull().sum() to find duplicate records sum

run command cars_data.drop_duplicates(inplace=True) to drop duplicate records

run command cars_data.shape to check the latest rows, column count

Now we have the clean data without null and duplicate

Perform Data Analysis

run command cars_data.info()

to see the datatypes

Check for unique records

run command

for col in cars_data.columns:

print(‘Unique values of ‘ + col )

print( cars_data[col].unique())

print(“=========\n” )

Write Function to Get Brand Name & Function to Convert to Float

def get_brand_name(car_name):
car_name = car_name.split(‘ ‘)[0]
return car_name.strip()

def clean_data(value):

value=value.split(‘ ‘)[0]
value=value.strip()
if value==”:
value=0
return float(value)

Check if brand name is getting extracted from car name by using the written function

def get_brand_name(car_name):

car_name = car_name.split(‘ ‘)[0]
return car_name.strip()

def clean_data(value):

value=value.split(‘ ‘)[0]
value=value.strip()
if value==”:
value=0
return float(value)

Now clean the other columns, do required typecasting

run follwing command

cars_data[‘mileage’] = cars_data[‘mileage’].apply(clean_data)

cars_data[‘max_power’] = cars_data[‘max_power’].apply(clean_data)

cars_data[‘engine’] = cars_data[‘engine’].apply(clean_data)

Check for unique car names

run follwing command

for col in cars_data.columns:
print(‘Unique values of ‘ + col )
print( cars_data[col].unique())
print(“=========\n” )

Assign numeric value to car names

run follwing command

cars_data[‘name’].replace([‘Maruti’, ‘Skoda’, ‘Honda’, ‘Hyundai’, ‘Toyota’, ‘Ford’, ‘Renault’, ‘Mahindra’,
‘Tata’, ‘Chevrolet’, ‘Datsun’, ‘Jeep’, ‘Mercedes-Benz’, ‘Mitsubishi’, ‘Audi’,
‘Volkswagen’, ‘BMW’, ‘Nissan’, ‘Lexus’, ‘Jaguar’, ‘Land’, ‘MG’, ‘Volvo’, ‘Daewoo’,
‘Kia’, ‘Fiat’, ‘Force’, ‘Ambassador’, ‘Ashok’, ‘Isuzu’, ‘Opel’],
[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]
,inplace=True)

It will run successfuly, but show this warning error, re run the above command again this error will disappear

Assign numeric value to car names

run follwing command

cars_data[‘transmission’].unique()

cars_data[‘transmission’].replace([‘Manual’, ‘Automatic’],[1,2],inplace=True)

cars_data[‘seller_type’].unique()

cars_data[‘seller_type’].replace([‘Individual’, ‘Dealer’, ‘Trustmark Dealer’],[1,2,3],inplace=True)

cars_data[‘fuel’].unique()

cars_data[‘fuel’].replace([‘Diesel’, ‘Petrol’, ‘LPG’, ‘CNG’],[1,2,3,4],inplace=True)

cars_data[‘owner’].unique()

cars_data[‘owner’].replace([‘First Owner’, ‘Second Owner’, ‘Third Owner’,
‘Fourth & Above Owner’, ‘Test Drive Car’],[1,2,3,4,5],inplace=True)

Rechecking Data Types

run follwing command

cars_data.info()

Check Data Set

run following command

cars_data

We now have all numerical values

Split Input and Output Data Set

run following command

input_data = cars_data.drop(columns=[‘selling_price’])
output_data = cars_data[‘selling_price’]

Split Train and Test Data Set

20% goes to test data set, 80% data to train data set

x_train, x_test, y_train, y_test = train_test_split(input_data, output_data, test_size=0.2)

Step 7 - Creating & Training Machine Learning Model

LinearRegression Model Creation

run following command

model = LinearRegression()

model.fit(x_train,y_train)

predict = model.predict(x_test)

Step 8 - Testing the outcome

Import pandas

run following command

import pandas as pd

Set the input values

input_data_model = pd.DataFrame(
[[5, 2018, 5000, 2, 1, 1, 1, 17.5, 1273, 100.1, 4]], # Data inside a list of lists
columns=[‘name’, ‘year’, ‘km_driven’, ‘fuel’, ‘seller_type’, ‘transmission’,
‘owner’, ‘mileage’, ‘engine’, ‘max_power’, ‘seats’] # Ensure correct list closing
)

Print and Check Input Value

print(input_data_model)

Run and Test The Outcome

model.predict(input_data_model)

Saving the model

run following command

import pickle as pk

pk.dump(model,open(‘model.pk1′,’wb’))

The model will be created and saved in same location as csv file location

The Above Was Building and Deploying ML model

Now to have a frontend UI - Download Visual Studio Code

Download for Windows

https://code.visualstudio.com/

Run the installer

code for app.py

in same folder place model and csv file

import pandas as pd

import numpy as np

import pickle as pk

import streamlit as st

model=pk.load(open(‘model.pk1′,’rb’))

st.header(‘ Car Price Prediction Model ‘)

cars_data = pd.read_csv(‘Cardetails.csv’)

def get_brand_name(car_name):

car_name = car_name.split(‘ ‘)[0]

return car_name.strip()

cars_data[‘name’] = cars_data[‘name’].apply(get_brand_name)

name = st.selectbox(‘Select Car Brand’, cars_data[‘name’].unique())

year = st.slider(‘Car Manufactured Year’, 1994,2024)

km_driven = st.slider(‘No of kms Driven’, 11,200000)

fuel = st.selectbox(‘Fuel type’, cars_data[‘fuel’].unique())

seller_type = st.selectbox(‘Seller type’, cars_data[‘seller_type’].unique())

transmission = st.selectbox(‘Transmission type’, cars_data[‘transmission’].unique())

owner = st.selectbox(‘Seller type’, cars_data[‘owner’].unique())

mileage = st.slider(‘Car Mileage’, 10,40)

engine = st.slider(‘Engine CC’, 700,5000)

max_power = st.slider(‘Max Power’, 0,200)

seats = st.slider(‘No of Seats’, 5,10)

if st.button(“Predict”):

input_data_model = pd.DataFrame(

[[name,year,km_driven,fuel,seller_type,transmission,owner,mileage,engine,max_power,seats]],

columns=[‘name’,’year’,’km_driven’,’fuel’,’seller_type’,’transmission’,’owner’,’mileage’,’engine’,’max_power’,’seats’])

input_data_model[‘owner’].replace([‘First Owner’, ‘Second Owner’, ‘Third Owner’,

‘Fourth & Above Owner’, ‘Test Drive Car’],

[1,2,3,4,5], inplace=True)

input_data_model[‘fuel’].replace([‘Diesel’, ‘Petrol’, ‘LPG’, ‘CNG’],[1,2,3,4], inplace=True)

input_data_model[‘seller_type’].replace([‘Individual’, ‘Dealer’, ‘Trustmark Dealer’],[1,2,3], inplace=True)

input_data_model[‘transmission’].replace([‘Manual’, ‘Automatic’],[1,2], inplace=True)

input_data_model[‘name’].replace([‘Maruti’, ‘Skoda’, ‘Honda’, ‘Hyundai’, ‘Toyota’, ‘Ford’, ‘Renault’,

‘Mahindra’, ‘Tata’, ‘Chevrolet’, ‘Datsun’, ‘Jeep’, ‘Mercedes-Benz’,

‘Mitsubishi’, ‘Audi’, ‘Volkswagen’, ‘BMW’, ‘Nissan’, ‘Lexus’,

‘Jaguar’, ‘Land’, ‘MG’, ‘Volvo’, ‘Daewoo’, ‘Kia’, ‘Fiat’, ‘Force’,

‘Ambassador’, ‘Ashok’, ‘Isuzu’, ‘Opel’],

[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]

,inplace=True)

car_price = model.predict(input_data_model)

#st.markdown(‘Car Price is going to be : ‘+ str(car_price[0]))

st.markdown(‘Car Price is going to be : ‘ + str(abs(int(round(car_price[0])))))

Isstall and Run Streamlit in Terminal