In [28]:

    

%pylab inline
import pandas as pd
from sklearn.decomposition import PCA
from sklearn import preprocessing
from sklearn import linear_model


    

    




Populating the interactive namespace from numpy and matplotlib

In [29]:

    

train = pd.read_csv('train.csv')
new_labels = train.columns.values
new_labels[-1] = 'total_rentals'
train.columns = new_labels
train[:5]

print train.ix[:,9:12].shape

fig,ax = plt.subplots(1,3)
ax[0].plot(train.ix[:,9])
ax[1].plot(train.ix[:,10])
ax[2].plot(train.ix[:,11])
plt.show()


    

    




(10886, 3)






    

In [30]:

    

X = train.ix[:,1:9]
hours = [datetime.datetime.strptime(d,'%Y-%m-%d %H:%M:%S').hour for d in train.datetime]
X.insert(0,'time',hours)
y = train.ix[:,9:12]
print X.shape, y.shape

pca2 = PCA(n_components=2)
X_2 = pca2.fit_transform(X)
print X_2.shape


    

    




(10886, 9) (10886, 3)
(10886, 2)

In [31]:

    

test_set = pd.read_csv('test.csv')
hours = [datetime.datetime.strptime(d,'%Y-%m-%d %H:%M:%S').hour for d in test_set.datetime]
test_set.insert(1,'time',hours)
print test_set.shape
test_set[:5]


    

    




(6493, 10)






    
Out[31]:






  

    

      

      
datetime
      
time
      
season
      
holiday
      
workingday
      
weather
      
temp
      
atemp
      
humidity
      
windspeed
    
  
  

    

      
0
      
 2011-01-20 00:00:00
      
 0
      
 1
      
 0
      
 1
      
 1
      
 10.66
      
 11.365
      
 56
      
 26.0027
    
    

      
1
      
 2011-01-20 01:00:00
      
 1
      
 1
      
 0
      
 1
      
 1
      
 10.66
      
 13.635
      
 56
      
  0.0000
    
    

      
2
      
 2011-01-20 02:00:00
      
 2
      
 1
      
 0
      
 1
      
 1
      
 10.66
      
 13.635
      
 56
      
  0.0000
    
    

      
3
      
 2011-01-20 03:00:00
      
 3
      
 1
      
 0
      
 1
      
 1
      
 10.66
      
 12.880
      
 56
      
 11.0014
    
    

      
4
      
 2011-01-20 04:00:00
      
 4
      
 1
      
 0
      
 1
      
 1
      
 10.66
      
 12.880
      
 56
      
 11.0014
    
  

5 rows × 10 columns

In [32]:

    

lm = linear_model.LinearRegression()
lm.fit(X,y)
print lm.score(X,y)
y_ = lm.predict(test_set.ix[:,1:])
fig,ax = plt.subplots(1,4)
for i,d in enumerate(y_.T):
    ax[i].plot(d)
ax[3].plot(y.ix[:,0]+y.ix[:,1])


    

    




0.316606687211






    
Out[32]:





[<matplotlib.lines.Line2D at 0x116ac6fd0>]






    

In [26]:

    

sample_submission = pd.read_csv('sampleSubmission.csv')
new_labels = sample_submission.columns.values
new_labels[-1] = 'total_rentals'
sample_submission.columns = new_labels
sample_submission[:5]

my_submission = sample_submission.copy()
my_submission.total_rentals = np.round(y_[:,2])
my_submission.loc[my_submission.total_rentals < 0,'total_rentals'] = 0
plt.plot(my_submission.total_rentals)
plt.show()
print my_submission.shape
my_submission[:5]


    

    













    




(6493, 2)






    
Out[26]:






  

    

      

      
datetime
      
total_rentals
    
  
  

    

      
0
      
 2011-01-20 00:00:00
      
 15
    
    

      
1
      
 2011-01-20 01:00:00
      
 23
    
    

      
2
      
 2011-01-20 02:00:00
      
 30
    
    

      
3
      
 2011-01-20 03:00:00
      
 39
    
    

      
4
      
 2011-01-20 04:00:00
      
 46
    
  

5 rows × 2 columns

In [27]:

    

new_labels = my_submission.columns.values
new_labels[-1] = 'count'
my_submission.columns = new_labels
my_submission.to_csv('my_submission.csv',index=False)


    

In [ ]: