POC CO2 emission.ipynb

Kernel: Python 3

Technical Problem Statement

Climate change due to carbon dioxide (CO ₂) emissions is one of the most complex challenges threatening our planet. This issue considered as a great and international Concern that primary attributed from different fossil fuels.

• To reduce the effect of Global Warming caused mainly due to cars(CO ₂) manufacturing companies are pre-estimating the CO₂ emissions for a newly manufactured car that is about to launch in the near future based on the features of the car.

• This project deals with estimating the CO₂ Emissions for a newly manufactured car by making Regression models which can accurately predetermine the car CO₂ emissions before it is being launched.

We have a sample dataset of year 2014 manufactured cars with their brands & other important specifications.lets see how we can design predetermine the co2 emision of a car.

About Data Set

The Data given here is for the Year 2014 manufactured Cars.

• YEAR – Year of manufacturing of car.

• MAKE – Manufacturing company name.

• VEHICLECLASS – Type of vehicles like SUV or medium-sized etc.

• ENGINESIZE – Size of the car’s engine (expressed in cc or cubic centimetre).

• CYLINDERS – Number of Cylinders in the engine.

• TRANSMISSION – Automatic or manual transmission with the number of gears.

• FUELTYPE – It indicates the type of fuel car use i.e. Diesel, Petrol, Z (Unleaded Petrol) etc.

• FUELCONSUMPTION_CITY – Fuel consumption or Fuel economy of car while running in city expressed in miles per gallon.

• FUELCONSUMPTION_HWY - Fuel Consumption or Fuel economy of car on highway expressed in miles per gallon.

• FUELCONSUMPTION_COMB – Net or combination of Fuel Economy expressed in miles per gallon.

• FUELCONSUMPTION_COMB_MPG – Total fuel economy expressed in miles per gallon.

• CO2EMISSIONS – The CO₂ emitted by the car expressed in grams.

Importing Data Set

import pandas as pd
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error, r2_score
import matplotlib.pyplot as plt
import pandas_profiling

c=pd.read_excel("F:\ML & Data Visualization\CO2 emission.xlsx")

c.head()

# Data Profiling
DataPre_report = pandas_profiling.ProfileReport(c)
DataPre_report.to_file(outputfile="myoutputfile.html")

## Identify the target variable 
c['FUELTYPE'],a = pd.factorize(c['FUELTYPE'])
c['VEHICLECLASS'],a = pd.factorize(c['VEHICLECLASS'])
c['TRANSMISSION'],a = pd.factorize(c['TRANSMISSION'])
c['MODEL'],a = pd.factorize(c['MODEL'])

c.head()

# Dropping feilds MODEL YEAR & CYLINDERS
c = c.drop(['MODELYEAR','CYLINDERS','MAKE'], axis=1)
X = c.drop('CO2EMISSIONS', axis=1)
y = c[['CO2EMISSIONS']]

## we’ll split the dataset into a train set and a test set. 
## Scikit-learn has a very straightforward train_test_split function for that.

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2,
                                                  random_state=1)

Lets build the regression model. First, let’s try a model with only one variable

#CO2 EMISSION VERSES FUEL TYPE
reg = LinearRegression()
reg.fit(X_train[['ENGINESIZE']], y_train)

LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None,
         normalize=False)

# Evaluation metrics for linear regression are mean squared error and the R² score.
y_predicted = reg.predict(X_test[['ENGINESIZE']])
print("Mean squared error: %.2f" % mean_squared_error(y_test, y_predicted))
print('R²: %.2f' % r2_score(y_test, y_predicted))

Mean squared error: 955.37
R²: 0.76

except e as exception 
try : 

Insights

• The best possible score is 1.0, We get a model with a mean squared error of 955.37 and an R² of0.76.

reg = LinearRegression()
reg.fit(X_train[['ENGINESIZE','FUELCONSUMPTION_CITY','FUELCONSUMPTION_COMB']], y_train)
y_predicted = reg.predict(X_test[['ENGINESIZE','FUELCONSUMPTION_CITY','FUELCONSUMPTION_COMB']])
print("Mean squared error: %.2f" % mean_squared_error(y_test, y_predicted))
print('R²: %.2f' % r2_score(y_test, y_predicted))

Mean squared error: 590.67
R²: 0.85