Get some data to play with



In [ ]:

    
from sklearn.datasets import load_digits
digits = load_digits()
digits.keys()



In [ ]:

    
digits.images.shape



In [ ]:

    
print(digits.images[0])



In [ ]:

    
import matplotlib.pyplot as plt
%matplotlib notebook

plt.matshow(digits.images[0], cmap=plt.cm.Greys)



In [ ]:

    
digits.data.shape



In [ ]:

    
digits.target.shape



In [ ]:

    
digits.target

Data is always a numpy array (or sparse matrix) of shape (n_samples, n_features)

Split the data to get going



In [ ]:

    
from sklearn.cross_validation import train_test_split
X_train, X_test, y_train, y_test = train_test_split(digits.data,
                                                    digits.target)

Really Simple API

0) Import your model class



In [ ]:

    
from sklearn.svm import LinearSVC

1) Instantiate an object and set the parameters



In [ ]:

    
svm = LinearSVC(C=0.1)

2) Fit the model



In [ ]:

    
svm.fit(X_train, y_train)

3) Apply / evaluate



In [ ]:

    
print(svm.predict(X_train))
print(y_train)



In [ ]:

    
svm.score(X_train, y_train)



In [ ]:

    
svm.score(X_test, y_test)

And again



In [ ]:

    
from sklearn.ensemble import RandomForestClassifier



In [ ]:

    
rf = RandomForestClassifier(n_estimators=50)



In [ ]:

    
rf.fit(X_train, y_train)



In [ ]:

    
rf.score(X_test, y_test)



In [ ]: