motorcortex-python-tools

Command-line tools and a Python helper library for logging and plotting signals from a Motorcortex server.

This package ships:

• mcx-datalogger — subscribe to server signals and stream them to a CSV (or LZMA-compressed .csv.xz) file, optionally gated by a trigger condition.
• mcx-dataplot — plot signals from a logged CSV file with flexible multi-axis / multi-subplot layouts.
• motorcortex_tools — a small Python package with the DataLogger class, a loadData() helper, and a waitFor() utility, for building your own logging and automated-testing scripts.

Installation

Linux (system Python + apt)

sudo apt install python3 python3-matplotlib python3-jinja2
sudo pip3 install motorcortex-python motorcortex-python-tools

Any platform (pip)

pip3 install matplotlib jinja2 motorcortex-python motorcortex-python-tools

Requires Python 3.x and motorcortex-python. pandas, matplotlib and jinja2 are pulled in as dependencies. The optional autotest extra (pip install motorcortex-python-tools[autotest]) adds weasyprint for PDF report generation.

Full tool help:

mcx-datalogger --help
mcx-dataplot --help

mcx-datalogger

Quick start

# Log a set of signals until Ctrl-C
mcx-datalogger -p parameters.json -u wss://192.168.2.100:5568:5567 -s mcx.cert.crt

# Use a trigger: start/stop logging when a boolean signal goes high/low
mcx-datalogger -p parameters.json \
               --trigger root/ControlSystem/logEnable \
               --triggervalue 1 --triggerop ==

# Write compressed output into a folder, with a comment in the filename
mcx-datalogger -p parameters.json -F ./logs -c bench_run_42 -C

Output filenames default to <UTC timestamp>[_comment].csv (or .csv.xz with -C). A sibling <file>.params dump of all server parameters is written alongside each trace unless --noparamdump is set.

Parameter file (-p/--parameterfile)

Two JSON formats are accepted.

Legacy — flat list of paths:

[
  {"path": "root/Control_task/actual_cycle_max"},
  {"path": "root/Fieldbus_task/actual_cycle_max"}
]

Scope export — object with a subscriptions array:

{
  "subscriptions": [
    {"path": "root/MainControlLoop/goToMoves/velocityFilter/output",
     "channel": 0, "color": "#109618", "axis": "y1", "hidden": false},
    {"path": "root/MainControlLoop/goToMoves/velocityFilter/output",
     "channel": 1},
    {"path": "root/MainControlLoop/jointPositionsTarget",
     "channel": 0}
  ]
}

This is the format exported by the updated Motorcortex scope / plot tool. Extra fields (channel, color, axis, hidden, …) are ignored for logging, and repeated paths are de-duplicated — a signal with multiple channels is logged once, not once per channel.

Key options

mcx-dataplot

# List the columns in a logged file
mcx-dataplot -l data.csv

# Plot signals 1 and 2 on the same axis; 3 on a second y-axis; 4 and 5 in a new subplot
mcx-dataplot -s 1,2:3 4,5 data.csv

# Save the plot to disk instead of showing it
mcx-dataplot -s 1,2 --output plot.png data.csv

-s/--signals syntax:

• , — add signal to the same axis
• : — put signal on an additional y-axis within the same subplot
• space (between -s arguments) — start a new subplot

Compressed .xz inputs are handled transparently.

motorcortex_tools (library)

from motorcortex_tools import DataLogger, loadData, waitFor

logger = DataLogger(
    "wss://192.168.2.100:5568:5567",
    ["root/signal1", "root/signal2"],
    divider=10, certificate="mcx.cert.crt",
)
logger.openFileAndWriteHeader("trace.csv")
logger.start()
# ... wait for some condition ...
logger.stop()
logger.close()

# Later, from any Python process:
df = loadData("trace.csv")          # pandas.DataFrame
df_with_time = loadData("trace.csv.xz", nodateconv=False)

DataLogger can also log to memory — call start() without openFileAndWriteHeader() first and read from logger.traces.

See the automatic_testing_examples/ folder for a fuller scripted-test pattern (uses waitFor() with comparison operators).

Changelog

Notable changes are tracked in CHANGELOG.md.

Release process (maintainers)

1. Update CHANGELOG.md — move items out of [Unreleased] into a new version section dated today.
2. Bump the version in setup.py.
3. Tag the release:

   git tag -a <version> -m "version <version>"
   git push origin <version>
   

4. Build:

   python3 ./setup.py sdist bdist_wheel
   

5. Upload to PyPI:

   python3 -m twine upload dist/motorcortex-python-tools-<version>.tar.gz
   

License

MIT — see LICENSE. © 2019–2026 VECTIONEER.