Converting Raw Accelerometer Data to Activity Counts Using Open-Source Code: Implementing a MATLAB Code in Python and R, and Comparing the Results to ActiLife
Bibliographic record
Abstract
Background: Closed-source software for processing and analyzing accelerometer data provides little to no information about the algorithms used to transform acceleration data into physical activity indicators. Recently, an algorithm was developed in MATLAB that replicates the frequently used proprietary ActiLife activity counts. The aim of this software profile was (a) to translate the MATLAB algorithm into R and Python and (b) to test the accuracy of the algorithm on free-living data. Methods: As part of the INTErventions, Research, and Action in Cities Team, data were collected from 86 participants in Victoria (Canada). The participants were asked to wear an integrated global positioning system and accelerometer sensor (SenseDoc) for 10 days on the right hip. Raw accelerometer data were processed in ActiLife, MATLAB, R, and Python and compared using Pearson correlation, interclass correlation, and visual inspection. Results: Data were collected for a combined 749 valid days (>10 hr wear time). MATLAB, Python, and R counts per minute on the vertical axis had Pearson correlations with the ActiLife counts per minute of .998, .998, and .999, respectively. All three algorithms overestimated ActiLife counts per minute, some by up to 2.8%. Conclusions: A MATLAB algorithm for deriving ActiLife counts was implemented in R and Python. The different implementations provide similar results to ActiLife counts produced in the closed source software and can, for all practical purposes, be used interchangeably. This opens up possibilities to comparing studies using similar accelerometers from different suppliers, and to using free, open-source software.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.037 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.017 | 0.013 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".