Classification Rank Budget
Rank complete labeled scores and inspect every top-K review budget with exact prefix counts.
- 1Add input
- 2Adjust settings
- 3Get your result
Tool input and files are processed in this browser without being uploaded.
Before you start
Choose a positive class, score direction and tie policy, then see how a fixed review budget changes precision, recall and F1 across the entire held-out set.
How to use this tool
- Paste complete CSV or choose one file, then select that active source.
- Set distinct one-based ID, label and score columns, positive/negative labels, direction and tie policy.
- Run, inspect any K, and save all six files with the full copied report.
Supported inputs and limits
Strict UTF-8 CSV with one header and 1–1,000,000 complete records; unique nonempty string IDs, two explicit labels and finite decimal scores.
64 MiB source, 10,000,000 source cells, 12,000,000 curve cells and 100,000,000 combined work units; limits apply together.
Files plus full text ≤512 MiB, typed result ≤512 MiB, their aggregate ≤768 MiB, wire and owned memory ≤1 GiB. A legal record count alone does not guarantee joint capacity.
One 120-second deadline covers capture, read, validation, calculation, encoding, cleanup and first publication; a limit failure returns no partial curve.
Score ties preserve every K. No positives gives undefined recall and F1=0 for every K. Preview is limited to 16 budgets and 4,000 codepoints; downloads and full copy are complete.
Worked example
Example input
id,label,score a,1,1 b,0,-0 c,1,0
Example options
Synthetic example; positive=1, negative=0, higher first, source-order ties.
Example output
K=1: TP=1, precision=1, recall=1/2, F1=2/3; K=3 retains all rows, TP=2, F1=4/5.
When something does not work
Changing input or settings cancels the run and clears old downloads. After cancellation or timeout, retry the same original in a fresh worker; fix malformed input or reduce a budget-exceeding scope first.
Frequently asked questions
Why do tied scores still create separate budget rows?
A review budget selects exactly K records. Equal scores use the chosen secondary order and original source ordinal; score thresholds would merge several budgets.
Can I use the original issue’s 2000-row data?
The public issue includes a plot, but no original sample values. The starter and validation fixtures are explicitly synthetic.