self.name = name
self.start, self.stop = start, stop
self.ls = None
def __call__(self, nn, train_history):
if self.ls is None:
self.ls = np.linspace(self.start, self.stop, nn.max_epochs)
epoch = train_history[-1]['epoch']
new_value = float32(self.ls[epoch - 1])
getattr(nn, self.name).set_value(new_value)
if __name__ == '__main__':
train_size = 0.75
cls = RandomForestClassifier()
train_df_orig = RevenueCompetition.load_data()
y = train_df_orig['revenue'].values.astype('float32')
del train_df_orig['revenue']
test_df_orig = RevenueCompetition.load_data(train=False)
full_df_orig = train_df_orig.append(test_df_orig)
print("Transforming...")
tr = make_pipeline(RevenueTransform(rescale=True), StandardScaler())
tr.fit(full_df_orig)
print('Classify the outliers...')
ly = np.log(y)
ym = ly.mean()
ys = ly.std()
from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier
from sklearn.cross_validation import StratifiedShuffleSplit
from sklearn.svm import SVR
from revenue import RevenueCompetition, RevenueTransform
if __name__ == '__main__':
train_size = 0.75
cls = RandomForestClassifier()
reg = RandomForestRegressor(n_estimators=20, max_features=5, max_depth=None,
min_samples_split=2, min_samples_leaf=1,
max_leaf_nodes=None, bootstrap=True,
oob_score=False, n_jobs=-1)
reg = SVR(C=10., gamma=0.1)
train_df_orig = RevenueCompetition.load_data()
y = train_df_orig['revenue'].values
del train_df_orig['revenue']
test_df_orig = RevenueCompetition.load_data(train=False)
full_df = train_df_orig.append(test_df_orig)
print("Transforming...")
tr = RevenueTransform(rescale=True)
tr.fit(full_df)
X = tr.transform(train_df_orig).values
print('Classify the outliers...')
ly = np.log(y)
ym = ly.mean()
