Adding a dumbbell plot which shows activities as line segments showing start->end times by time of day and day of year.
It includes an option for the user to specify a local timezone, which is a temporary solution. It would be better for all time-related pre-processing to be done in process_data as timezone specifications also impact the calendar plot and future time-related plots.
Ideally the timezone could be inferred from lat / lon in the GPX / FIT files on a point-by-point basis (to address the unlikely edge case where the timezone changes during an activity), perhaps using https://pypi.org/project/tzwhere/, then joined to activities using the "Filename" column in activities.csv.
Other issues:
Activities that cover multiple days (e.g. 10pm Day 1 to 3am Day 2) are not being handled correctly - need to be split up into separate line segments.
There is a daylight savings time issue which is yet to be resolved - temp solution is to use .dt.tz_localize(tz='UTC', nonexistent='NaT', ambiguous='NaT') which can result in data loss.
Adding a dumbbell plot which shows activities as line segments showing start->end times by time of day and day of year.
It includes an option for the user to specify a local timezone, which is a temporary solution. It would be better for all time-related pre-processing to be done in process_data as timezone specifications also impact the calendar plot and future time-related plots.
Ideally the timezone could be inferred from lat / lon in the GPX / FIT files on a point-by-point basis (to address the unlikely edge case where the timezone changes during an activity), perhaps using https://pypi.org/project/tzwhere/, then joined to activities using the "Filename" column in activities.csv.
Other issues: