Skip to content

README.md

Latest commit

 

History

History
42 lines (39 loc) · 3.41 KB

README.md

File metadata and controls

42 lines (39 loc) · 3.41 KB

Running EmotiBit Brainflow SpO2 Algorithm using pybind

Requirements

  • Python
  • Python virtual environment (refer to the Emotibit Plugins repo for specific instructions on how to setup)
  • CMake
  • EmotiBit ArduinoFilters
    • The EmotiBit ArduinoFilter directory should be on the same level as this repository.
    • Option 1: Library Manager
      1. Open Arduino IDE,
      2. open the Library Manager
      3. Download EmotiBit_ArduinoFilters v1.0.1 (if not already installed)
    • Option 2: git clone (example showing cloning in Arduino Libraries directory, but user can clone in any other location, as long as ArduinoFilters is also cloned at the same directory level)
      1. cd into Arduino libraries folder
        • Windows: C:\Users\{username}\Documents\Arduino\libraries
        • Mac: /Users/{username}/Documents/Arduino/libraries
        • Linux: /home/{username}/Arduino/libraries
      2. Run git clone https://github.com/EmotiBit/EmotiBit_ArduinoFilters

Build

  • Open a terminal window
  • cd into the pybind directory
  • Before being able to run cmake, you will need to update the path to pybind11 (if required) in the CMakeLists.txt file. pybind11 is installed in the python virtual environment, as the link in the previous section explains.
    • We tested this with virtual environment created using python 3.12, and therefore the path to pybind in the current CMakeLists.txt file is py-env/lib/python3.12/site-packages/pybind11/share/cmake/pybind11/.
  • Run cmake -B build to generate the build files (the build files will be under a directory named build)
  • Run cmake --build build --config Release to generate the .pyd file
    • On Windows using MSVC, the .pyd file will be under pybind/build/Release
    • Otherwise, the .pyd file will be under pybind/build
  • Activate the Python virtual environment (instructions found in Emotibit Plugins repo)

Running run.py

  • Open run.py and uncomment lines 6-7 (if on linux/macOS), or uncomment lines 10-11 (if on Windows)
  • run.py runs on emotibit data. We have checked in test data in thie repositoty. By default, the script looks for a data directory on the same level. Therefore, to test the script, you can extract the tests/sit-stand-sit_v0.0.0/ip900ap/sit-stand-sit_v0.0.0_ip900ap.zip file and place the EMotiBit.csv file in the data folder. You can then use the EmotiBit data parser to parse the files, since the script looks for files ending in _PR and _PI.

Testing

  • Unzip the .zip file in tests/simulated-unobstructed-airway_v0.0.0/ip900ap
  • Run the EmotiBit Data Parser on tests/simulated-unobstructed-airway_v0.0.0/ip900ap/EmotiBit.csv
  • Verify that EmotiBit_PI.csv, EmotiBit_PR.csv, and EmotiBit_UN.csv exist under tests/simulated-unobstructed-airway_v0.0.0/ip900ap
  • Switch to the terminal window in the pybind directory
  • Run python run.py ../tests/simulated-unobstructed-airway_v0.0.0/ip900ap
    • matplotlib will need to be installed for the script to run successfully
  • A plot of the raw PPG data overlayed with the calculated SpO2 should appear Note: The timestamp used by run.py is the EmotiBitTimestamp column in the Emotibit .csv files