def random_forest(X_train, X_test, y_train, y_test):
'''
Fits the random forest determined by grid search to the training data, returns various scores.
INPUT: The train test split.
'''
rf_model = RandomForestClassifier(bootstrap=False,
class_weight=None,
criterion='gini',
max_depth=None,
max_features=4,
max_leaf_nodes=None,
min_samples_leaf=9,
min_samples_split=9,
min_weight_fraction_leaf=0.0,
n_estimators=20,
n_jobs=1,
oob_score=False,
random_state=456,
verbose=0,
warm_start=False)
rf_model.fit(X_train, y_train)
y_pred = rf_model.predict(X_test)
print score_models(y_test, y_pred)
def baseline_classifier(y_test):
'''
Scores a "stupid" model that just predicts the majority class with probability equal to the proportion of happy-ending movies.
'''
y_pred_base_odds = []
for i in range(len(y_test)):
num = np.random.choice(np.arange(0,2), p=[(1-.535), .535])
y_pred_base_odds.append(num)
print score_models(y_test, y_pred_base_odds)
def majority_classifier(y_test):
'''
Scores a "stupid" model that just predicts the majority class every time (except once to avoid divide by zero error.)
'''
y_pred_always_happy = [1]*(len(y_test))
print score_models(y_test, y_pred_always_happy)