Brainflow

Brainflow is a library intended to obtain, parse and analyze EEG\EMG\ECG data.

Core module of this library is implemented in C\C++ and available for all bindings as a dynamic library.

Build status

Linux(Travis):

Windows(AppVeyour):

Shared library methods:

int prepare_session (int board_id, const char *port_name);
int start_stream (int buffer_size);
int stop_stream ();
int release_session ();
int get_current_board_data (int num_samples, float *data_buf, double *ts_buf, int *returned_samples);
int get_board_data_count (int *result);
int get_board_data (int data_count, float *data_buf, double *ts_buf);
int set_log_level (int log_level);

For now only OpenBCI Cyton is supported, but architecture allows to add new boards and not only from OpenBCI without any issues

Bindings

I support bindings for:

These bindings just call methods from dynamic libraries

Python sample

import argparse
import time
import brainflow


def main ():
    parser = argparse.ArgumentParser ()
    parser.add_argument ('--port', type = str, help  = 'port name', required = True)
    parser.add_argument ('--log', action = 'store_true')
    args = parser.parse_args ()

    if (args.log):
        brainflow.board_shim.BoardShim.enable_board_logger ()
    else:
        brainflow.board_shim.BoardShim.disable_board_logger ()

    board = brainflow.board_shim.BoardShim (brainflow.board_shim.CYTON.board_id, args.port)
    board.prepare_session ()
    board.start_stream ()
    time.sleep (5)
    data = board.get_board_data ()
    board.stop_stream ()
    board.release_session ()

    data_handler = brainflow.preprocess.DataHandler (brainflow.board_shim.CYTON.board_id, numpy_data = data)
    filtered_data = data_handler.preprocess_data (order = 3, start = 1, stop = 50)
    data_handler.save_csv ('results.csv')
    print (filtered_data.head ())
    read_data = brainflow.preprocess.DataHandler (brainflow.board_shim.CYTON.board_id, csv_file = 'results.csv')
    print (read_data.get_data ().head ())


if __name__ == "__main__":
    main ()

Brainflow Emulator

Brainflow emulator allows users to run tests without EEG headset.

On Windows it works using com0com
On Linux it works using pty

You should pass command line to test directly to cyton_linux.py or to cyton_windows.py, script will add port automatically to provided command line and will start an application

Example

Install emulator package

cd emulator
pip3 install -U .

Run test

python3 brainflow_emulator/cyton_linux.py python3 /media/sf_folder/brainflow/python-package/examples/brainflow_get_data.py --log --port
INFO:root:Running python3 /media/sf_folder/brainflow/python-package/examples/brainflow_get_data.py --log --port /dev/pts/1
INFO:root:read "b'v'"
INFO:root:read "b'b'"
INFO:root:read "b's'"
INFO:root:stdout is: b'   package_num         eeg1          eeg2  ...    accel2    accel3     timestamp\n0          0.0    59.692543    264.193723  ...  3.526750  1.616875  1.553124e+09\n1          1.0   425.466525   2773.043094  ...  1.519125 -3.823000  1.553124e+09\n2          2.0  1349.970005  12973.129874  ...  2.586875  0.657750  1.553124e+09\n3          3.0  2730.721543  35263.509116  ... -2.776625  1.778750  1.553124e+09\n4          4.0  4333.067974  60181.588357  ...  3.000500 -0.270625  1.553124e+09\n\n[5 rows x 13 columns]\n   package_num         eeg1          eeg2  ...    accel2    accel3     timestamp\n0          0.0    59.692543    264.193723  ...  3.526750  1.616875  1.553124e+09\n1          1.0   425.466525   2773.043094  ...  1.519125 -3.823000  1.553124e+09\n2          2.0  1349.970005  12973.129874  ...  2.586875  0.657750  1.553124e+09\n3          3.0  2730.721543  35263.509116  ... -2.776625  1.778750  1.553124e+09\n4          4.0  4333.067974  60181.588357  ...  3.000500 -0.270625  1.553124e+09\n\n[5 rows x 13 columns]\n'
INFO:root:stderr is: b'[2019-03-20 19:12:10.313] [board_logger] [info] openning port /dev/pts/1: 0\n'

Emulator raises TestFailureError if test fails

Brainflow Tests

Instructions to run tests

You need to install emulator package first

cd {project_dir}/emulator
python setup.py install

Run Python tests

cd {project_dir}/python-package
pip install -e .
# to test it on Linux machine
python3 {project_dir}/emulator/brainflow_emulator/cyton_linux.py python3 {project_dir}/tests/python/brainflow_get_data.py
# to test it on Windows machine
python3 {project_dir}/emulator/brainflow_emulator/cyton_windows.py python3 {project_dir}/tests/python/brainflow_get_data.py

Run Java test

cd {project_dir}\java-package
# I dont add jar to repo, so you will need to build it
mvn package
cd {project_dir}/tests/java
javac -classpath {project_dir}\java-package\brainflow\target\brainflow-java-jar-with-dependencies.jar BrainFlowTest.java
python {project_dir}\emulator\brainflow_emulator\cyton_windows.py java -classpath .;{project_dir}\java-package\brainflow\target\brainflow-java-jar-with-dependencies.jar BrainFlowTest

Build instructions

Windows

Install CMAKE>=3.10
Run cmake_build.cmd You may change this line to use VS you have installed

Linux

Install CMAKE>=3.10
Run bash cmake_build.sh

If you wanna distribute compiled Linux linraries you HAVE to build it inside this docker continer otherwise your libraries will fail on Linux with another glibc version

hack-r / brainflow