Examples
Examples
Try the starting example
- Supplied ranking relevance in displayed order
- 3,2,0,1
- Cutoff k (whole number)
- 3
- Gain convention: exponential or linear
- exponential
- NDCG at supplied cutoff
- 0.946768
- Discounted cumulative gain
- 8.892789
- Ideal DCG within supplied candidate pool
- 9.392789
- Effective cutoff
- 3
Try another scenario
- Supplied ranking relevance in displayed order
- 3,2,0,1
- Cutoff k (whole number)
- 3
- Gain convention: exponential or linear
- linear
- NDCG at supplied cutoff
- 0.894999
- Discounted cumulative gain
- 4.26186
- Ideal DCG within supplied candidate pool
- 4.76186
- Effective cutoff
- 3
Calculation method
Calculation method
NDCG at supplied cutoff
DCG/ideal DCG; explicitly 0 when ideal gain is 0.
Discounted cumulative gain
Sum gain(relevance)/log2(rank+1), starting at rank 1.
Ideal DCG within supplied candidate pool
Sort only supplied relevance values descending before applying the same gain and cutoff.
Effective cutoff
Minimum of supplied k and candidate count.
Symbols and units
- a
- Supplied ranking relevance in displayed order
- b
- Cutoff k (whole number)
- c
- Gain convention: exponential or linear
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
- Ranking relevance must be supplied; no relevance judgments are inferred. Ideal ranking is restricted to the supplied pool, gain is explicit, and zero ideal gain uses an explicit NDCG convention of 0.
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.