def test_connection():
"""Test TCP/IP connection for StimServer <-> StimClient.
"""
global _server, _have_put_in_trigger
# have to start a thread to simulate the effect of two
# different computers since stim_server.start() is designed to
# be a blocking method
# use separate queues because timing matters
trig_queue1 = queue.Queue()
trig_queue2 = queue.Queue()
# start a thread to emulate 1st client
thread1 = threading.Thread(target=_connect_client, args=(trig_queue1, ))
thread1.daemon = True
# start another thread to emulate 2nd client
thread2 = threading.Thread(target=_connect_client, args=(trig_queue2, ))
thread2.daemon = True
thread1.start()
thread2.start()
with StimServer(port=4218, n_clients=2) as stim_server:
_server = stim_server
stim_server.start(timeout=10.0) # don't allow test to hang
# Add the trigger to the queue for both clients
stim_server.add_trigger(20)
_have_put_in_trigger = True # monkey patch
# the assert_equal must be in the test_connection() method
# Hence communication between threads is necessary
trig1 = trig_queue1.get(timeout=_max_wait)
trig2 = trig_queue2.get(timeout=_max_wait)
assert_equal(trig1, 20)
# test if both clients receive the same trigger
assert_equal(trig1, trig2)
# test timeout for stim_server
with StimServer(port=4218) as stim_server:
assert_raises(StopIteration, stim_server.start, 0.1)
data_path = sample.data_path()
raw_fname = data_path + '/MEG/sample/sample_audvis_filt-0-40_raw.fif'
raw = mne.io.read_raw_fif(raw_fname, preload=True)
fig, ax = plt.subplots(1)
ax.set(xlabel='Trials',
ylabel='Classification score (% correct)',
title='Real-time feedback')
isi = 0.01 # this is unrealistic, but will make the example run quickly
n_trials = 40 # number of trials to simulate
n_start = 5 # number of trials to run before decoding
rng = np.random.RandomState(0)
# Instantiating stimulation server
# The with statement is necessary to ensure a clean exit
with StimServer(port=4218) as stim_server:
# The channels to be used while decoding
picks = mne.pick_types(raw.info, meg='grad')
rt_client = MockRtClient(raw)
# Constructing the pipeline for classification
# don't highpass filter because of short signal length of epochs
scaler = preprocessing.StandardScaler()
vectorizer = Vectorizer()
clf = SVC(C=1, kernel='linear')
concat_classifier = Pipeline([('vector', vectorizer), ('scaler', scaler),
('svm', clf)])
ev_list = list(rng.randint(3, 5, n_start)) # some random starting events
score_lv, score_rv, score_x = [], [], []
from mne.datasets import sample
from mne.realtime import StimServer
from mne.realtime import MockRtClient
from mne.decoding import ConcatenateChannels, FilterEstimator
print(__doc__)
# Load fiff file to simulate data
data_path = sample.data_path()
raw_fname = data_path + '/MEG/sample/sample_audvis_filt-0-40_raw.fif'
raw = mne.io.Raw(raw_fname, preload=True)
# Instantiating stimulation server
# The with statement is necessary to ensure a clean exit
with StimServer('localhost', port=4218) as stim_server:
# The channels to be used while decoding
picks = mne.pick_types(raw.info,
meg='grad',
eeg=False,
eog=True,
stim=True,
exclude=raw.info['bads'])
rt_client = MockRtClient(raw)
# Constructing the pipeline for classification
filt = FilterEstimator(raw.info, 1, 40)
scaler = preprocessing.StandardScaler()
concatenator = ConcatenateChannels()
