Untitled

import pandas as pd

def download_data():
    import os
    import urllib

    if os.path.exists('cars.csv'):
        print('Data already downloaded')
        return

    urllib.request.urlretrieve('https://think.cs.vt.edu/corgis/datasets/csv/cars/cars.csv','cars.csv')
    print('Download Complete')

download_data()

df = pd.read_csv('cars.csv')

df.head()

df = df[df['Make'] == 'Honda']
df = df[df['Fuel Information.Fuel Types'] == 'Gasoline']
df = df[df['Fuel Information.Combined MPG'] < df['Fuel Information.Combined MPG'].quantile(0.9)]

features = ['Fuel Information.Highway MPG', 'Engine Information.Engine Statistics.Horsepower']

x = df[features]

x = x.values

x = x.astype('float32')

x = (x-x.min())/(x.max()-x.min())

y = df['Fuel Information.Combined MPG']

y = y.astype('float32')

y = (y-y.min())/(y.max()-y.min())

y = y.values

import matplotlib.pyplot as plt

plt.scatter(x[:,0], y, c='r',label='data')

plt.show()

plt.scatter(x[:,1], y, c='r',label='data')

plt.show()

from keras.models import Sequential

from keras.layers import Dense

model = Sequential()

model.add(Dense(1, input_dim=2, kernel_initializer='normal', activation='linear'))

model.compile(loss='mse', optimizer='adam', metrics=['mse'])

model.fit(x, y, epochs=100, batch_size=1, verbose=1)

scores = model.evaluate(x, y, verbose=0)

print(model.get_weights())