Examples
Examples
Try the starting example
- JSON Classes [ID,true-positive,false-positive,false-negative]
- [["A",80,10,20],["B",20,20,10]]
- Class F1 and support
- [["A",0.842105263158,100],["B",0.571428571429,30]]
- Micro F1
- 0.769231
- Macro F1
- 0.706767
- Support-weighted F1
- 0.779641
Try another scenario
- JSON Classes [ID,true-positive,false-positive,false-negative]
- [["A",70,20,30],["B",25,15,5]]
- Class F1 and support
- [["A",0.736842105263,100],["B",0.714285714286,30]]
- Micro F1
- 0.730769
- Macro F1
- 0.725564
- Support-weighted F1
- 0.731637
Calculation method
Calculation method
Class F1 and support
F1=2TP/(2TP+FP+FN), support=TP+FN; undefined class F1 returns zero.
Micro F1
Compute F1 after summing TP, FP, FN across classes.
Macro F1
Unweighted mean of per-class F1 including zero-support classes.
Support-weighted F1
Mean per-class F1 weighted by TP+FN; all-zero support returns zero.
Symbols and units
- a
- JSON Classes [ID,true-positive,false-positive,false-negative]
Use the units shown. A list keeps the order in which you enter its values. The symbol ^ means “raised to the power of”.
Assumptions & limits
Assumptions & limits
- Per-class one-vs-rest confusion-count F1 comparison. Counts must derive from one consistent dataset; this contract does not attempt to reconstruct unprovided confusion cells. Supported limits: All numeric inputs must be finite. Only the explicitly described mathematical scenario is supported.
Are these inputs private?
These calculations run in your browser. A shared link contains the values you choose to share. Only questions submitted to the separate AI finder are sent to OpenAI.
Further reading
Background reading; this does not imply endorsement or independent review of SumFizz.