Scripts to detect artifacts and in electrodermal activity (EDA) data. Note that these scripts are written in Python 2.7.
Please also cite this project: Taylor, S., Jaques, N., Chen, W., Fedor, S., Sano, A., & Picard, R. Automatic identification of artifacts in electrodermal activity data. In Engineering in Medicine and Biology Conference. 2015.
Required python packages:
- numpy: 1.9.2
- scipy: 0.14.0
- pandas: 0.16.0
- sklearn: 0.16.1
- matplotlib: 1.3.1
- PyWavelets: 0.2.2
To run artifact detection from the command line:
Note that PickleDirectory is the main directory
Currently there are only 2 classifiers to choose from: Binary or Multiclass
To run peak detection:
Descriptions of the algorithm settings can be found at http://eda-explorer.media.mit.edu/info/
To run accelerometer feature extraction:
This file works slightly differently than the others in that it gives summary information over periods of time.
Currently, these files are written with the assumption that the sample rate is an integer power of 2.
Keep the "classify.py" and "SVMBinary.p" and "SVMMulticlass.p" files in the same directory.
Please visit eda.explorer.media.mit.edu to use the web-based version