Skip to content

marcelblijleven/fittie

Repository files navigation

fittie

Parse Garmin .FIT files

PyPI version

Installation

Fittie is available on pypi and can be installed with the following command.

$ pip install fittie

Example

from fittie import decode

if __name__ == "__main__":
    fitfile = decode("path/to/fit/file.fit")
        
    # Example: get average heart rate
    print(fitfile.average_heart_rate)

    # Loop through all data messages:
    for data_message in fitfile:
        print(data_message)

For more information and examples, check the documentation

Fitfile

Usage

Decoding / parsing a FIT file is done through the decode function in the fittie.fitfile package. It accepts the following types of arguments:

  • A file path string
  • A file opened in "rb" mode
  • A buffered reader, BinaryIO or BytesIO
# Examples
from io import BytesIO
from fittie.fitfile import decode

fitfile_1 = decode("/path/to/fit/file.fit")         # Path to file

fitfile_2 = decode(BytesIO(...))                    # BytesIO

with open("/path/to/fit/file.fit", "rb") as f:      # File opened in rb mode
    fitfile_3 = decode(f)

To view the available message types in the fitfile, use the available_message_types property. It will return a list of message type keys. These keys can be used to retrieve all messages of a certain kind. After retrieving the available message types, the messages can be retrieved using get_messages_by_type.

fitfile = decode("/path/to/fit/file.fit")

types = fitfile.available_message_types
# e.g. [ 'file_id', 'device_info', 'record', 'event', 'lap', 'session', 'activity']
messages = fitfile.get_messages_by_type('record')  # Returns a list of `DataMessage`

Alternatively, you can interact with the messages property of fitfile directly, this is a simple dict.

File types

All FIT files should contain a file id message that describes the type of file. Common file types are activity, workout and course. More file types can be found in fit_types.py.

To retrieve the type of the decoded fitfile, use the .file_type property.

assert fitfile.file_type == "activity"

CRC

A crc check is done by default, but can be disabled by providing calculate_crc=False to the decode function to improve speed.

For example, on the same FIT file with 58297 data messages, decoding with crc takes 0.029 seconds and without crc it only takes 0.014 seconds.

DataMessages

To access data in a DataMessage, use the fields property. This will return a dict with all the values inside the message.

fitfile = decode("/path/to/fit/file.fit")

for record in fitfile.get_messages_by_type("record")[:5]:
    print(record.fields)

# {'timestamp': 1044776016}
# {'timestamp': 1044776016, 'heart_rate': 117}
# {'timestamp': 1044776017, 'heart_rate': 116}
# {'timestamp': 1044776017, 'heart_rate': 115}
# {'timestamp': 1044776018, 'heart_rate': 115}

TODO:

  • Handle component fields
  • Handle accumulators
  • Handle chained FIT files
  • Handle compressed timestamps
  • move record_header into record, instead of reading it separately
  • encoding