def __init__(self,
user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics):
self.module_id = str(uuid.uuid4())
self.user_id = user_id
self.device_type = device_type
self.rmq_address = rmq_address
self.rmq_user = rmq_user
self.rmq_pwd = rmq_pwd
self.input_metrics = None
self.output_metrics = None
self.eeg_subscriber = None
self.mu_publisher = None
self.routing_keys = None
self.preprocessor = None
self.num_channels = get_num_channels(self.device_type, "eeg")
self.eeg_data = np.zeros((0, self.num_channels))
self.count = 0
self.eyeblinks_remover = EyeBlinksFilter()
self.step_size = None
self.electrodes_placement = None
self.enable_ica = False
self.started_fit = False
def train(self):
categories = range(self.num_categories)
num_channels = get_num_channels(self.device_type, self.input_metric)
labels = []
input_values = []
for category in categories:
input_file = TRAINING_DATA_PATH % {
"tag": category,
"metric": self.input_metric,
"device": self.device_type
}
with open(input_file, "rb") as csvFile:
reader = csv.reader(csvFile)
headers = reader.next()
for row in reader:
data = dict(zip(headers, row))
channel_values = []
category = data["tag"]
for i in range(num_channels):
channel = "channel_%s" % i
channel_values.append(float(data[channel]))
input_values.append(channel_values)
labels.append(category)
num_records = len(input_values)
num_train_records = num_records * 2 / 3
train_input_values = input_values[:num_train_records]
test_input_values = input_values[num_train_records:]
train_labels = labels[:num_train_records]
test_labels = labels[num_train_records:]
X = np.array(train_input_values)
y = np.array(train_labels)
if self.classifier_type == "svm":
clf = svm.LinearSVC()
else:
raise ValueError("Classifier type is '%s' but can only be: %s"
% VALID_CLASSIFIER_TYPES)
clf.fit(X, y)
X_test = np.array(test_input_values)
y_test = np.array(test_labels)
score = clf.score(X_test, y_test)
classifier_path = CLASSIFIER_PATH % {
"classifier_type": self.classifier_type,
"device": self.device_type,
"input_metric": self.input_metric
}
model_path = os.path.join(
os.path.dirname(brainsquared.__file__), "module_runners",
classifier_path)
joblib.dump(clf, model_path)
return classifier_path, score
def __init__(self,
user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id=None):
"""
Collect, tag data (optional) and write to a CSV file.
:param user_id: ID of the user using the device.
:type user_id: string
:param device_type: type of the device publishing to this module.
:type device_type: string
:param rmq_address: address of the RabbitMQ server.
:type rmq_address: string
:param rmq_user: login for RabbitMQ connection.
:type rmq_user: string
:param rmq_pwd: password for RabbitMQ connection.
:type rmq_pwd: string
:param input_metrics: name of the input metric.
:type input_metrics: dict
:param output_metrics: name of the output metric.
:type output_metrics: dict
:param module_id: (Optional. Default = None) ID of the module
:type module_id: string
"""
super(CSVWriterSink, self).__init__(user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id)
self.num_input_channels = get_num_channels(device_type, self.input_metric)
# Set when configure() is called.
self.tag = None
self.time_offset = None
self.recording_time = None
self.csv_writer = None
# Module specific.
self.start_time = int(time.time() * 1000)
self.channels = ["channel_%s" % i for i in range(self.num_input_channels)]
self.headers = ["timestamp"] + self.channels
def __init__(self,
user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id=None):
"""
Classify with simple thresholds.
:param user_id: ID of the user using the device.
:type user_id: string
:param device_type: type of the device publishing to this module.
:type device_type: string
:param rmq_address: address of the RabbitMQ server.
:type rmq_address: string
:param rmq_user: login for RabbitMQ connection.
:type rmq_user: string
:param rmq_pwd: password for RabbitMQ connection.
:type rmq_pwd: string
:param input_metrics: name of the input metric.
:type input_metrics: dict
:param output_metrics: name of the output metric.
:type output_metrics: dict
:param module_id: (Optional. Default = None) ID of the module
:type module_id: string
"""
super(ThresholdClassifier, self).__init__(user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id)
self.num_input_channels = get_num_channels(device_type, self.input_metric)
# Set when configure() is called.
self.thresholds = None
# Module specific. Keep track of the last classification.
self._last_classification = {
"timestamp": int(time.time() * 1000),
"channel_0": self.num_input_channels
}
def __init__(self,
user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id=None):
"""
:param user_id: ID of the user using the device.
:type user_id: string
:param device_type: type of the device publishing to this module.
:type device_type: string
:param rmq_address: address of the RabbitMQ server.
:type rmq_address: string
:param rmq_user: login for RabbitMQ connection.
:type rmq_user: string
:param rmq_pwd: password for RabbitMQ connection.
:type rmq_pwd: string
:param input_metrics: name of the input metric.
:type input_metrics: dict
:param output_metrics: name of the output metric.
:type output_metrics: dict
:param module_id: (Optional. Default = None) ID of the module
:type module_id: string
"""
super(SKLearnClassifier, self).__init__(user_id,
device_type,
rmq_address,
rmq_user,
rmq_pwd,
input_metrics,
output_metrics,
module_id)
self.num_input_channels = get_num_channels(device_type, self.input_metric)
# Set when configure() is called.
self.classifier_path = None
self.classifier = None
self.classifier_type = None
self.num_categories = None