CPT data cleaning workflow

Why it matters

Cone penetration test data often arrives with inconsistent headers, unit drift, and depth gaps that make early interpretation harder than it should be.

Workflow

1. Import the raw CPT export without altering column names.
2. Validate units, depth ordering, and null runs before plotting.
3. Record assumptions beside the cleaned dataset so downstream readers can reproduce the choices.

Output expectations

• A cleaned tabular dataset with preserved provenance.
• A first-pass qc plot for tip resistance, sleeve friction, and pore pressure.
• Notes that explain where automated cleaning stopped and engineering judgement began.

Limitations

This seeded notebook is a starter read path, not a substitute for project-specific review, calibration, or site interpretation.