[docs] Add doctests for metrics

[ydcjeff]

📚 Documentation

Following #2230 , I listed the metrics places (suggested by @sdesrozis) where doctests can be added.

Metrics

• Average ([skip ci] add doctest for accumulation metrics #2332)
• GeometricAverage ([skip ci] add doctest for accumulation metrics #2332)
• VariableAccumulation (This class is an helper, no test needed)
• Accuracy ([skip ci] add doctest for Accuracy metric #2345)
• ConfusionMatrix ([skip ci] add doctest for confusion matrix metrics #2336)
• ClassificationReport ([skip ci] Add doctest for ClassificationReport metric #2346)
• DiceCoefficient ([skip ci] add doctest for confusion matrix metrics #2336)
• JaccardIndex ([skip ci] add doctest for confusion matrix metrics #2336)
• IoU ([skip ci] add doctest for confusion matrix metrics #2336)
• mIoU ([skip ci] add doctest for confusion matrix metrics #2336)
• EpochMetric ([skip ci] Add doctest for EpochMetric #2351)
• Fbeta ([skip ci] Add doctest for Precision, Recall and FBeta metrics #2340)
• Loss ([skip ci] Doctest for Loss #2335)
• MeanAbsoluteError ([skip ci] Doctests for MeanAbsoluteError and MeanSquaredError #2280)
• MeanPairwiseDistance (Doctests for RMSE and MeanPairwiseDistance #2307)
• MeanSquaredError ([skip ci] Doctests for MeanAbsoluteError and MeanSquaredError #2280)
• MetricsLambda ([skip ci] Add doctest for MetricsLambda #2352)
• MultiLabelConfusionMatrix ([skip ci] Add doctest for MultiLabelConfusionMatrix metric #2347)
• Precision ([skip ci] Add doctest for Precision, Recall and FBeta metrics #2340)
• PSNR (Doctest for PSNR #2311)
• Recall ([skip ci] Add doctest for Precision, Recall and FBeta metrics #2340)
• RootMeanSquaredError (Doctests for RMSE and MeanPairwiseDistance #2307)
• RunningAverage ([skip ci] Add doctest for RunningAverage #2348)
• SSIM (feat(doctest): SSIM metric doctest #2241)
• TopKCategoricalAccuracy ([skip ci] Doctest for TopKCategoricalAccuracy #2284)
• Bleu (Doctest for nlp metrics (BLEU, Rouge, RougeN, RougeL) #2317)
• Rouge (Doctest for nlp metrics (BLEU, Rouge, RougeN, RougeL) #2317)
• RougeL (Doctest for nlp metrics (BLEU, Rouge, RougeN, RougeL) #2317)
• RougeN (Doctest for nlp metrics (BLEU, Rouge, RougeN, RougeL) #2317)
• InceptionScore ([skip ci] Add doctest for GAN metrics #2349)
• FID ([skip ci] Add doctest for GAN metrics #2349)

Contrib Metrics

• AveragePrecision ([skipci] [Doctest] added contrib metrics AveragePrecision, PrecisionRecallCurve, ROC_AUC and RocCurve #2341)
• CohenKappa ([skip ci] add doctest for CohenKappa metric #2321)
• PrecisionRecallCurve ([skipci] [Doctest] added contrib metrics AveragePrecision, PrecisionRecallCurve, ROC_AUC and RocCurve #2341)
• ROC_AUC ([skipci] [Doctest] added contrib metrics AveragePrecision, PrecisionRecallCurve, ROC_AUC and RocCurve #2341)
• RocCurve ([skipci] [Doctest] added contrib metrics AveragePrecision, PrecisionRecallCurve, ROC_AUC and RocCurve #2341)
• CanberraMetric ([skipci] [Doctest] added contrib regression (FractionalBias, FractionalAbsoluteError, CanberraMetric, GeometricMeanAbsoluteError) #2323)
• FractionalAbsoluteError ([skipci] [Doctest] added contrib regression (FractionalBias, FractionalAbsoluteError, CanberraMetric, GeometricMeanAbsoluteError) #2323)
• FractionalBias ([skipci] [Doctest] added contrib regression (FractionalBias, FractionalAbsoluteError, CanberraMetric, GeometricMeanAbsoluteError) #2323)
• GeometricMeanAbsoluteError ([skipci] [Doctest] added contrib regression (FractionalBias, FractionalAbsoluteError, CanberraMetric, GeometricMeanAbsoluteError) #2323)
• GeometricMeanRelativeAbsoluteError ([skip ci] add doctest for regression metrics #2324)
• ManhattanDistance ([skip ci] add doctest for regression metrics #2324)
• MaximumAbsoluteError ([skip ci] add doctest for regression metrics #2324)
• MeanAbsoluteRelativeError ([skip ci] add doctest for regression metrics #2324)
• MeanError ([skip ci] add doctest for regression metrics #2324)
• MeanNormalizedBias ([skip ci] add doctest for regression metrics #2324)
• MedianAbsoluteError ([skip ci] add doctest for regression metrics #2324)
• MedianAbsolutePercentageError ([skip ci] add doctest for regression metrics #2324)
• MedianRelativeAbsoluteError ([skip ci] add doctest for regression metrics #2324)
• R2Score ([skip ci] add doctest for regression metrics #2324)
• WaveHedgesDistance ([skip ci] add doctest for regression metrics #2324)

This is a great issue for first comers and those who want to learn more about Sphinx. This issue is the community effort.
If you want to contribute, please comment on this issue with the metric you want to add doctest (one by one).

Example on how to add doctest, see #2241

Thanks!

[wanjuntham]

Hello,

May I take up Average?

[vfdev-5]

Sure @wanjuntham , go for it, thanks!

[halsawadi]

Hi, everyone

I'm coming from PyData Global 2021's sprint.

I'll work on Precision, if it isn't taken yet.

[Eunjnnn]

Hello
This is my first contribution :-)

can I work on Loss ?

[sdesrozis]

Hello

This is my first contribution :-)

can I work on Loss ?

@Eunjnnn thank you ! Yes you can 😉

[wanjuntham]

@vfdev-5 Hey, I actually didnt have time to work on this. Could you unassign me?

[sdesrozis]

@Eunjnnn Are you still working on Loss ? Or do you need help ?

[Eunjnnn]

@sdesrozis Hi I'm trying to work on it, but actually i'm not sure where to start,,

[sdesrozis]

@sdesrozis Hi I'm trying to work on it, but actually i'm not sure where to start,,

First, take a look to our contribution guide. Then, you will have to set up your environment (testing, building the doc).

And finally, you will find here what is needed to write the doc and doctests according to our guidelines :)

[Ishan-Kumar2]

I'll add some of the contrib regression metrics FractionalBias, FractionalAbsoluteError, CanberraMetric, GeometricMeanAbsoluteError

[Ishan-Kumar2]

I'll add the remaining contrib metrics AveragePrecision, PrecisionRecallCurve, ROC_AUC and RocCurve