pampro is a software project for the systematic analysis of physical activity data collected in epidemiological studies. The ultimate goal is to provide a turn-key solution for physical activity data analysis, replicating published methodologies in a monitor-agnostic framework.

The main feature of this version is a HDF5 module, which provides functions to store pampro objects inside HDF5 containers. This has numerous advantages, the most important being incredibly fast loading times for triaxial Time Series data (<10 seconds for a week long 100 Hz file). It also provides a mechanism to automatically cache the results of various time consuming functions, such as nonwear detection and autocalibration; this means if a second analysis is performed, the results will be instantly recalled from storage. In the future, it will become a low-memory solution for pampro analyses, which will minimise the RAM footprint by writing results directly to disk. A demonstration of how to use the new format in pampro can be found here.

All other changes are optimisations made on under-the-hood code, either to improve speed or memory efficiency, that have no impact on how users will write their analyses.

• Import channels of time series data from a variety of common monitors:
  • ActiHeart (.txt)
  • Axivity binary (.cwa)
  • GeneActiv (.bin)
  • Actigraph (.dat)
  • activPAL & activPAL micro binary (.datx)
  • Any timestamped data (.csv)
• Output piecewise summary statistics of any data channel, over any size time window:
  • Time spent in any cutpoint.
  • Sum, mean, percentiles, min, max.
• Visualise the time series data.
• Extract bouts of activity in any cutpoint.
• Various triaxial acceleration methodologies:
  • Nonwear detection.
  • Autocalibration.

Click here for a walkthrough of pampro's most basic features. Please note that designing an analysis currently requires extensive knowledge of the Python programming language. See /examples for example scripts demonstrating various more advanced features. The growing /methods section provides detailed explanations of the methods implemented in pampro, linking to the relevant literature where appropriate.

In your terminal, navigate to the desired installation directory and enter the following:

git clone https://github.com/Thomite/pampro.git
git checkout tags/v0.4.0
cd pampro
ipython setup.py install

This will clone the repository, sync to the latest official release (v0.4.0) and run the Python script to install it on your system. This presupposes that Git and Python are installed already.

If you use pampro in your work, it would be appreciated if you could cite it with the appropriate DOI: