Transform row C with the train-only min and max.

highlighted = computed this step

Apply the train fit to row C

Row C keeps the train-only fit. Its scaled value is 20 minus 0 over 10.

20010=2{20 - 0\over10}=2
Train Only Preprocessing ExactlyExact min-max fit with train rows and full-data contrast.role tablerowrolexAtrain0Btrain10Ctest20fit parametersfitminmaxtrain-only A,B010full-data A,B,C contrast020train-only min-max fit and full-data contrastrowxtrain-only scaledfull-data scaledA000B1011/2C2021test row C changes the full-data fit; contrast onlyfit preprocessing on train rows only; full-data is contrasttest row changes numbers when it enters the fitdoes not claim: NOT leakage guarantee; NOT future performanceNOT accuracy; NOT generalization; NOT calibrated; NOT safeNOT best preprocessing; NOT probability truth

C becomes two

The train-only scaled value for C is exactly 2.

Ctrain-fit=2C_{\text{train-fit}}=2
Train Only Preprocessing ExactlyExact min-max fit with train rows and full-data contrast.role tablerowrolexAtrain0Btrain10Ctest20fit parametersfitminmaxtrain-only A,B010full-data A,B,C contrast020train-only min-max fit and full-data contrastrowxtrain-only scaledfull-data scaledA000B1011/2C2021test row C changes the full-data fit; contrast onlyfit preprocessing on train rows only; full-data is contrasttest row changes numbers when it enters the fitdoes not claim: NOT leakage guarantee; NOT future performanceNOT accuracy; NOT generalization; NOT calibrated; NOT safeNOT best preprocessing; NOT probability truth

Do not refit on C

C is transformed by the fitted train-only numbers. It does not change those fitted numbers.

fit first; apply to test row\text{fit first; apply to test row}
Train Only Preprocessing ExactlyExact min-max fit with train rows and full-data contrast.role tablerowrolexAtrain0Btrain10Ctest20fit parametersfitminmaxtrain-only A,B010full-data A,B,C contrast020train-only min-max fit and full-data contrastrowxtrain-only scaledfull-data scaledA000B1011/2C2021test row C changes the full-data fit; contrast onlyfit preprocessing on train rows only; full-data is contrasttest row changes numbers when it enters the fitdoes not claim: NOT leakage guarantee; NOT future performanceNOT accuracy; NOT generalization; NOT calibrated; NOT safeNOT best preprocessing; NOT probability truth