sciforge.papertrail

Lineage Visualiser

Track the flow from raw data to final claims and detect when results go stale.

Dataset
cleaned_results.csv
Transform
remove_outliers()
Figure
fig2_roc_curve.png
Claim
AUC > 0.92 achieved

Python API

with sf.papertrail.session("v3") as trail:
    df = trail.load("data.csv")
    df2 = trail.transform(df, fn)
    fig = trail.figure("fig1", df2)
    trail.claim("AUC > 0.9", fig)

# Later:
report = sf.papertrail.check_stale("v3")
print(report.summary())
Papertrail stores content hashes of every node. If data.csv is modified, all downstream figures and claims are automatically flagged as STALE.
← Back to examples