A Linear Regression Example from the book "Principles of Data Science"



In [2]:

    
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
from sklearn import metrics
%matplotlib inline



In [3]:

    
url = 'https://raw.githubusercontent.com/justmarkham/DAT8/master/data/bikeshare.csv'
bikes = pd.read_csv(url)
bikes.head()









    Out[3]:






  
    
      
      datetime
      season
      holiday
      workingday
      weather
      temp
      atemp
      humidity
      windspeed
      casual
      registered
      count
    
  
  
    
      0
      2011-01-01 00:00:00
      1
      0
      0
      1
      9.84
      14.395
      81
      0.0
      3
      13
      16
    
    
      1
      2011-01-01 01:00:00
      1
      0
      0
      1
      9.02
      13.635
      80
      0.0
      8
      32
      40
    
    
      2
      2011-01-01 02:00:00
      1
      0
      0
      1
      9.02
      13.635
      80
      0.0
      5
      27
      32
    
    
      3
      2011-01-01 03:00:00
      1
      0
      0
      1
      9.84
      14.395
      75
      0.0
      3
      10
      13
    
    
      4
      2011-01-01 04:00:00
      1
      0
      0
      1
      9.84
      14.395
      75
      0.0
      0
      1
      1



In [4]:

    
bikes[['count', 'temp']].corr()



In [5]:

    
from sklearn.linear_model import LinearRegression
linreg = LinearRegression()
linreg.fit(bikes[['temp']], bikes['count'])
print(linreg.intercept_)
print(linreg.coef_)
count_predict = linreg.predict(bikes[['temp']])
np.sqrt(metrics.mean_squared_error(bikes['count'], count_predict))









    



6.04621295962
[ 9.17054048]






    Out[5]:





166.44886243326746



In [6]:

    
features = ['temp', 'season', 'weather', 'humidity']
linreg = LinearRegression()
linreg.fit(bikes[features], bikes['count'])
linreg.coef_
sns.pairplot(bikes, x_vars=features, y_vars='count', kind='reg')
count_predict = linreg.predict(bikes[features])
np.sqrt(metrics.mean_squared_error(bikes['count'], count_predict))









    Out[6]:





155.99832684186401



In [7]:

    
from sklearn.model_selection import train_test_split
features = bikes[['temp']]
count = bikes['count']
features_train, features_test, count_train, count_test = train_test_split(features, count)
print("total feature size: {}, train feature size: {}, test feature size: {}".format(len(features), len(features_train), len(features_test)))
linreg = LinearRegression()
linreg.fit(features_train, count_train)
count_predict = linreg.predict(features_test)
np.sqrt(metrics.mean_squared_error(count_test, count_predict))









    



total feature size: 10886, train feature size: 8164, test feature size: 2722






    Out[7]:





167.52513660585501



In [8]:

    
count_avg = bikes['count'].mean()
null_model_count_predict = [count_avg] * len(bikes)
np.sqrt(metrics.mean_squared_error(count, null_model_count_predict))









    Out[8]:





181.1361335742659



In [31]:









    Out[31]:





10886



In [20]:



In [ ]:

	datetime	season	weather	temp	atemp	humidity	casual	registered	count
0	2011-01-01 00:00:00	1	1	9.84	14.395	81	3	13	16
1	2011-01-01 01:00:00	1	1	9.02	13.635	80	8	32	40
2	2011-01-01 02:00:00	1	1	9.02	13.635	80	5	27	32
3	2011-01-01 03:00:00	1	1	9.84	14.395	75	3	10	13
4	2011-01-01 04:00:00	1	1	9.84	14.395	75	0	1	1