In [3]:

    

%pylab inline
from sklearn import datasets
from sklearn.metrics import mean_squared_error
from sklearn import linear_model
from sklearn.ensemble import RandomForestClassifier
boston = datasets.load_boston() # Boston house-prices
X = boston['data']   # 13 features (e.g. crime, # rooms, age, etc.)
Y = boston['target'] # response (median house price)
half = floor(len(Y)/2)
train_X = X[:half]
train_Y = Y[:half]
test_X = X[half:]
test_Y = Y[half:]


    

    




Populating the interactive namespace from numpy and matplotlib

In [18]:

    

rf1 = RandomForestClassifier(n_estimators=1000, max_features=6)
rf1.fit(train_X,train_Y)
preds = rf1.predict(test_X)
mse = mean_squared_error(test_Y, preds); print mse


    

    




53.079486166

In [17]:

    

plot(test_Y, preds-test_Y, 'o')


    

    
Out[17]:





[<matplotlib.lines.Line2D at 0x113697d90>]






    

In [14]:

    

lr = linear_model.LinearRegression()


    

In [15]:

    

lr.fit(train_X,train_Y)
lr_preds = lr.predict(test_X)
lr_mse = mean_squared_error(test_Y, lr_preds); print lr_mse


    

    




302.64091133

In [21]:

    

from sklearn.cross_validation import cross_val_score
import sklearn.cross_validation as cross_v


    

In [23]:

    

scores = cross_val_score(rf1, X, Y, cv = cross_v.KFold(len(Y),5))


    

In [ ]: