Robust Scaler
Dataset T-Shirt
| SrNo | Height(in cms) | Weight(in kgs) | T Shirt Size |
|---|
| 1 | 158 | 58 | M |
| 2 | 158 | 59 | M |
| 3 | 158 | 63 | M |
| 4 | 160 | 59 | M |
| 5 | 160 | 60 | M |
| 6 | 163 | 61 | M |
| 7 | 160 | 64 | L |
| 8 | 163 | 64 | L |
| 9 | 165 | 61 | L |
| 10 | 165 | 62 | L |
| 11 | 165 | 65 | L |
| 12 | 168 | 62 | L |
| 13 | 168 | 63 | L |
| 14 | 168 | 66 | L |
| 15 | 170 | 63 | L |
| 16 | 170 | 64 | L |
| 17 | 170 | 68 | L |
Robust Scale formular
$$x_{scale}=\frac{x_i-q_1}{q_3-q_1}$$
WITH CTE AS(
SELECT
APPROX_QUANTILES(Height, 4) HeightQ,
APPROX_QUANTILES(Weight, 4) WeightQ,
FROM Shop.TShirt
)SELECT
(Height - HeightQ[OFFSET(1)]) / (HeightQ[OFFSET(3)] - HeightQ[OFFSET(1)]) AS HeightScale,
(Weight - WeightQ[OFFSET(1)]) / (WeightQ[OFFSET(3)] - WeightQ[OFFSET(1)]) AS WeightScale,
Size
FROM Shop.TShirt, CTE;Result
| Height Scale | Weight Scale | T Shirt Size |
|---|
| -0.25 | -0.4 | M |
| -0.25 | -0.2 | M |
| -0.25 | 0.6 | M |
| 0 | -0.2 | M |
| 0 | 0 | M |
| 0.375 | 0.2 | M |
| 0.0 | 0.8 | L |
| 0.375 | 0.8 | L |
| 0.625 | 0.2 | L |
| 0.625 | 0.4 | L |
| 0.625 | 1.0 | L |
| 1.0 | 0.4 | L |
| 1.0 | 0.6 | L |
| 1.0 | 1.2 | L |
| 1.25 | 0.6 | L |
| 1.25 | 0.8 | L |
| 1.25 | 1.6 | L |