Python analysis

Requires pymavlink to be installed. If you need to parse messages recently introduced to MARSH that are not part of pymavlink yet, you will need to generate the code to read them. The script update_mavlink.py in sim-nodes repository can automatically download and generate the code for you.

import mavlink_all as mavlink
from pymavlink import mavutil

log_file = "./data/YYYYMMDDTHHMMSS_comment.tlog"

# We'll output the data as CSV in this example
print("t,roll,pitch,yaw")

parameters: dict[str, float] = {}
mav = mavutil.mavlink_connection(log_file, dialect="all")
start_time: float | None = None
while message := mav.recv_msg():
    # Convert to relative time from first message for simplicity
    if start_time is None:
        start_time = message._timestamp
    t = message._timestamp - start_time + parsed_time

    if message.get_type() == "PARAM_VALUE":
        # Save parameters that were in this run
        parameters[message.param_id] = message.param_value
    elif message.get_type() == "SIM_STATE":
        # After checking the message type we know it has fields that we want
        print(f"{t},{message.roll},{message.pitch},{message.yaw}")

# Resolve problem with mavutil not using custom dialect
import pymavlink.dialects.v20.all as _file_pymavlink_all  # pyright:ignore[reportMissingTypeStubs]
import mavlink_all as _file_marsh_all
import shutil

# HACK: mavutil always loads the modules by name from pymavlink directory, have to replace
_ = shutil.copy(_file_marsh_all.__file__, _file_pymavlink_all.__file__)