chironpy 0.29 — FIT File Merge Support and 1Hz Resampling
chironpy 0.29 introduces methods for merging multiple FIT files into a single continuous workout, addressing a common workflow when sessions are split across separate recordings.
Merging Workouts
The new WorkoutData.merge_many() class method accepts a list of separately recorded workouts and combines them into a single continuous WorkoutData object. Workouts are sorted by start timestamp, overlapping timestamps favour the later workout, and time gaps are preserved as NaN rows after resampling when interpolate=False.
merged = WorkoutData.merge_many([warmup, workout, cooldown])For convenience, WorkoutData.merge() wraps merge_many() when merging a single workout into the current instance.
Both methods support drop_gaps=True to shift workouts back-to-back without gap rows, which condenses elapsed time. Note that performance metrics will not account for rest time in this mode, as the gap is removed from the time series.
The distance column is forward-filled across gap rows in merged workouts. Distance is cumulative and should not appear as NaN during rest periods.
Setting Start Time
WorkoutData.set_start_time() shifts the datetime index of a workout to a specified start time, recomputing the time column. This is useful when a file (such as a Strava activity stream) contains no absolute timestamp and needs to be aligned with other workouts before merging.
workout.set_start_time("2026-04-04 10:00:00")Documentation and Examples
Three new Strava activity-stream example files demonstrate a single session split across three recordings: strava_17497731832_warm_up.json, strava_17497955000_track_workout.json, and strava_17498168651_warm_down.json.
The documentation now includes docs/features/merge_workouts.ipynb, an interactive notebook demonstrating merge_many(), merge(), and set_start_time() with these example files. A separate docs/Example.ipynb walks through analysing a workout from the Osaka Marathon 2025.
For developers looking to contribute or fork the project, a new lab/ environment provides Jupyter notebooks for rapid prototyping and local development.
Build System Update
The pyproject.toml build backend has been updated from poetry.masonry.api to poetry.core.masonry.api to support editable installs via pip install -e ..
Upgrade
Update to chironpy 0.29 via pip:
pip install --upgrade chironpyAbout chironpy
chironpy is Chiron’s free and open source Python library for processing and analysing endurance activity data. It standardises inputs from FIT, GPX, TCX, and Strava into a unified structure with 1Hz time-series data. It provides tools like best-effort intervals, elevation and grade analysis, training zone distribution, power metrics (WAP, stress score, W/kg), training stress and rolling/resampling tools — some of the same tooling used to power Chiron’s analytics.
Open source is how we give back to the endurance sports data science community. If you work with running or cycling data in Python, check it out.