Skip to content


One of the main ideas behind Sweatpy is to offer a generic interface to data from different data sources. A big part of this is making sure that the data frames that are returned by the sweat.read_*() methods (read_fit(), read_tcx(), etc.) are similar. I.e. it should not matter if you are importing an activity from a .fit file or a .tcx file: Working with the data should be the same.

For the data frames returned by the read_*() methods, the column and data types are as follows:

Column name unit comments
index or datetime pandas.DatetimeIndex Be aware that not all activity files are sampled at 1Hz
heartrate beats per minute
speed meters/second
power Watt
latitude degrees
longitude degrees
left-right balance %
cadence rounds per minute
distance meter
elevation meter
temperature degrees Celsius
lap - lap counter, zero-indexed. Only supported for .fit and .tcx files.
session - session counter, zero-indexed. Triggered when an activity is (manually or automatically) paused or stopped and then resumed. Only supported for .fit files.