def test_hat_nba(test_path):
stream = HyperplaneGenerator(mag_change=0.001,
noise_percentage=0.1,
random_state=2)
stream.prepare_for_use()
learner = HAT(leaf_prediction='nba')
cnt = 0
max_samples = 5000
y_pred = array('i')
y_proba = []
wait_samples = 100
while cnt < max_samples:
X, y = stream.next_sample()
# Test every n samples
if (cnt % wait_samples == 0) and (cnt != 0):
y_pred.append(learner.predict(X)[0])
y_proba.append(learner.predict_proba(X)[0])
learner.partial_fit(X, y)
cnt += 1
expected_predictions = array('i', [
1, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1,
0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 0, 1, 1, 1,
0
])
assert np.alltrue(y_pred == expected_predictions)
test_file = os.path.join(test_path, 'test_hoeffding_adaptive_tree_nba.npy')
data = np.load(test_file)
assert np.allclose(y_proba, data)
expected_info = "HAT(binary_split=False, bootstrap_sampling=True, grace_period=200,\n" \
" leaf_prediction='nba', max_byte_size=33554432,\n" \
" memory_estimate_period=1000000, nb_threshold=0, no_preprune=False,\n" \
" nominal_attributes=None, remove_poor_atts=False, split_confidence=1e-07,\n" \
" split_criterion='info_gain', stop_mem_management=False, tie_threshold=0.05)"
assert learner.get_info() == expected_info
assert type(learner.predict(X)) == np.ndarray
assert type(learner.predict_proba(X)) == np.ndarray
def test_hat_nba(test_path):
stream = HyperplaneGenerator(mag_change=0.001,
noise_percentage=0.1,
random_state=2)
stream.prepare_for_use()
learner = HAT(leaf_prediction='nba')
cnt = 0
max_samples = 5000
y_pred = array('i')
y_proba = []
wait_samples = 100
while cnt < max_samples:
X, y = stream.next_sample()
# Test every n samples
if (cnt % wait_samples == 0) and (cnt != 0):
y_pred.append(learner.predict(X)[0])
y_proba.append(learner.predict_proba(X)[0])
learner.partial_fit(X, y)
cnt += 1
expected_predictions = array('i', [
0, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 1,
0
])
assert np.alltrue(y_pred == expected_predictions)
test_file = os.path.join(test_path, 'test_hoeffding_adaptive_tree_nba.npy')
data = np.load(test_file)
assert np.allclose(y_proba, data)
expected_info = 'HAT: max_byte_size: 33554432 - memory_estimate_period: 1000000 - grace_period: 200' \
' - split_criterion: info_gain - split_confidence: 1e-07 - tie_threshold: 0.05' \
' - binary_split: False - stop_mem_management: False - remove_poor_atts: False' \
' - no_pre_prune: False - leaf_prediction: nba - nb_threshold: 0' \
' - nominal_attributes: [] - '
assert learner.get_info() == expected_info
assert type(learner.predict(X)) == np.ndarray
assert type(learner.predict_proba(X)) == np.ndarray
@author: Dr. Sayan Putatunda
"""
""" Chapter 3 Codes """
import os
os.getcwd() # to see the current path of working directory
os.chdir('./Python codes')
### Creating a synthetic dataset
from skmultiflow.data import HyperplaneGenerator
import pandas as pd
import numpy as np
create = HyperplaneGenerator(random_state=888,
n_features=10,
noise_percentage=0)
create.prepare_for_use()
X, Y = create.next_sample(10000)
data = pd.DataFrame(np.hstack((X, np.array([Y]).T)))
# Cast the last column to int
data = data.astype({10: int})
data.shape
# output- (10000, 11)
# Store it in csv
data.to_csv('data_stream.csv', index=False)
# Applying Hoeffding Tree on the synthetic data stream
import logging
from GooweMSS import GooweMS
import random
logger = logging.getLogger()
logger.setLevel(logging.INFO)
# Prepare the data stream
streams = []
N_STREAMS = 10
instances_num = 10000
for i in range(N_STREAMS):
stream = HyperplaneGenerator(random_state=None,
n_features=10,
n_drift_features=2,
mag_change=0.1,
noise_percentage=0.0,
sigma_percentage=0.1)
streams.append(stream)
stream.prepare_for_use()
stream_t = HyperplaneGenerator(random_state=None,
n_features=10,
n_drift_features=2,
mag_change=0.1,
noise_percentage=0.0,
sigma_percentage=0.1)
stream_t = streams[0]
stream_t.prepare_for_use()
instances_counter = 0
run(agrawal, 'agrawal_gen', DATASE_SIZE)
sea = ConceptDriftStream(SEAGenerator(classification_function=1,
noise_percentage=0.13),
SEAGenerator(classification_function=2,
noise_percentage=0.13),
position=DATASE_SIZE / 2)
run(sea, 'sea_gen', DATASE_SIZE)
led = LEDGeneratorDrift(has_noise=True,
noise_percentage=0.28,
n_drift_features=4)
run(led, 'led_gen', DATASE_SIZE)
stagger = ConceptDriftStream(STAGGERGenerator(classification_function=1,
balance_classes=False),
STAGGERGenerator(classification_function=2,
balance_classes=False),
position=DATASE_SIZE / 2)
run(stagger, 'stagger_gen', DATASE_SIZE)
hyperplane = HyperplaneGenerator(noise_percentage=0.28,
n_features=10,
mag_change=0.25,
sigma_percentage=0.3,
n_drift_features=5)
run(hyperplane, 'hyperplane_gen', DATASE_SIZE)
rbf = RandomRBFGeneratorDrift(change_speed=0.4)
run(rbf, 'rbf_gen', DATASE_SIZE)