# illustrate per class analysis using scikit 
# sergey v 

from sklearn.metrics import precision_recall_curve
from sklearn.metrics import average_precision_score
from sklearn.ensemble import ExtraTreesClassifier


etc = ExtraTreesClassifier()
etc.fit(X_train, y_train);
y_pred = etc.predict(X_test);
print("tree test set score: {:.2f}".format(np.mean(y_pred == y_test)))
print("Actual labels:\n {}".format(y_test.values))
print("tree predictions:\n {}".format(y_pred))


# define # of classes
n_classes = 6;

# one hot encode y_test and y_pred
y_test_orig=y_test.values
y_test=np.zeros((len(y_test_orig),n_classes))
for ii in range(len(y_test_orig)):
    y_test[ii,y_test_orig[ii]-1]=1


y_pred_orig=y_pred
y_pred=np.zeros((len(y_pred_orig),n_classes))
for ii in range(len(y_pred_orig)):
    y_pred[ii,y_pred_orig[ii]-1]=1


Y_test = y_test;
y_score = y_pred;

# loop over each class
precision = dict()
recall = dict()
average_precision = dict()
for i in range(n_classes):
    precision[i], recall[i], _ = precision_recall_curve(Y_test[:, i],
                                                        y_score[:, i])
    average_precision[i] = average_precision_score(Y_test[:, i], y_score[:, i])

# A "micro-average": quantifying score on all classes jointly
precision["micro"], recall["micro"], _ = precision_recall_curve(Y_test.ravel(),
    y_score.ravel())
average_precision["micro"] = average_precision_score(Y_test, y_score,
                                                     average="micro")
print('Average precision score, micro-averaged over all classes: {0:0.2f}'
      .format(average_precision["micro"]))