Home › Documentation

Documentation

A complete reference for how Data Sensors actually behaves - not just what each button is called, but what happens underneath it: when a calculated signal recomputes, what an auto-tune search actually searches, and what does or doesn't get saved to your session file.

Getting started

Opening a measurement file, importing a CAN trace, and building your first strip.

System Requirements

Minimum and recommended RAM, disk space, and CPU, and why Model Competition needs more.

Strips, signals, and cursors

Axis grouping, drag-to-reorder, zoom/pan, and placing dual cursors.

Calculated signals

Writing mathjs formulas, how and when they evaluate, and what gets saved.

Virtual sensor training and auto-tuning

Random Forest, XGBoost, and Neural Network models, with automatic hyperparameter search.

Model Competition

Training several model families at once and ranking them on a held-out validation split.

Virtual Signals panel and importing models

Every trained model as a reusable signal, and re-importing a previously exported model.

Simulink model viewer

Viewing a .mdl/.slx block diagram read-only - no MATLAB or Simulink license needed.

Saving, exporting, and slicing

Session configuration files, PNG strip export, and time-sliced measurement export.

Licensing

Node-locked and floating licenses, and the 7-day trial.

Related: Measurement Viewer Virtual Sensor Training Model Competition Standalone Code Export Licensing All features