Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

StandardScaler

Transformer sklearn-compatible 🔧 Preprocessing

StandardScaler — zero-mean unit-variance standardisation. / StandardScaler — standardisation à moyenne nulle et variance unitaire.

⚡ Rust-native ✓ sklearn parity
Quick start — Python 
import seraplot as sp, numpy as np
X = np.random.randn(200, 4) * [3, 0.5, 10, 1]
scaler = sp.StandardScaler()
scaler.fit(X)
Xt = scaler.transform(X)
print(Xt.mean(0), Xt.std(0))
💡
EN — Drop-in replacement: sp.StandardScaler has the same API as sklearn.
FR — Remplacement direct : même API que sklearn, changez l'import.

API Reference

JSON function name

ml_standard_scaler — aliases: standard_scaler, standardize

Python class
sp.StandardScaler()
Constructor Parameters

No constructor parameters.

Returns

JSON with transformed matrix.

Algorithm

$$x' = \frac{x - \mu}{\sigma}$$

Example
import seraplot as sp, numpy as np
X = np.random.randn(200, 4) * [3, 0.5, 10, 1]
scaler = sp.StandardScaler()
scaler.fit(X)
Xt = scaler.transform(X)
print(Xt.mean(0), Xt.std(0))

Référence API

Nom de fonction JSON

ml_standard_scaler — alias : standard_scaler, standardize

Classe Python
sp.StandardScaler()
Paramètres du constructeur

Aucun paramètre de constructeur.

Retourne

JSON avec matrice transformed.

Algorithme

$$x' = \frac{x - \mu}{\sigma}$$

Exemple
import seraplot as sp, numpy as np
X = np.random.randn(200, 4) * [3, 0.5, 10, 1]
scaler = sp.StandardScaler()
scaler.fit(X)
Xt = scaler.transform(X)
print(Xt.mean(0), Xt.std(0))

MinmaxScaler

Transformer sklearn-compatible 🔧 Preprocessing

MinMaxScaler — scale features to [0, 1] range. / MinMaxScaler — mise à l'échelle des features dans l'intervalle [0, 1].

⚡ Rust-native ✓ sklearn parity
Quick start — Python 
import seraplot as sp, numpy as np
X = np.random.randn(200, 3) * 100
scaler = sp.MinMaxScaler()
Xt = scaler.fit_transform(X)
print(Xt.min(), Xt.max())
💡
EN — Drop-in replacement: sp.MinmaxScaler has the same API as sklearn.
FR — Remplacement direct : même API que sklearn, changez l'import.

API Reference

JSON function name

ml_minmax_scaler — aliases: minmax_scaler, min_max_scaler

Python class
sp.MinmaxScaler()
Constructor Parameters

No constructor parameters.

Returns

JSON with transformed matrix.

Algorithm

$$x' = \frac{x - x_{\min}}{x_{\max} - x_{\min}}$$

Example
import seraplot as sp, numpy as np
X = np.random.randn(200, 3) * 100
scaler = sp.MinMaxScaler()
Xt = scaler.fit_transform(X)
print(Xt.min(), Xt.max())

Référence API

Nom de fonction JSON

ml_minmax_scaler — alias : minmax_scaler, min_max_scaler

Classe Python
sp.MinmaxScaler()
Paramètres du constructeur

Aucun paramètre de constructeur.

Retourne

JSON avec matrice transformed.

Algorithme

$$x' = \frac{x - x_{\min}}{x_{\max} - x_{\min}}$$

Exemple
import seraplot as sp, numpy as np
X = np.random.randn(200, 3) * 100
scaler = sp.MinMaxScaler()
Xt = scaler.fit_transform(X)
print(Xt.min(), Xt.max())

RobustScaler

Transformer sklearn-compatible 🔧 Preprocessing

RobustScaler — scale using median and IQR, robust to outliers. / RobustScaler — mise à l'échelle par médiane et IQR, robuste aux valeurs aberrantes.

⚡ Rust-native ✓ sklearn parity
Quick start — Python 
import seraplot as sp, numpy as np
X = np.random.randn(300, 4)
X[0, :] = 100
scaler = sp.RobustScaler()
Xt = scaler.fit_transform(X)
print(Xt.mean(0))
💡
EN — Drop-in replacement: sp.RobustScaler has the same API as sklearn.
FR — Remplacement direct : même API que sklearn, changez l'import.

API Reference

JSON function name

ml_robust_scaler — aliases: robust_scaler

Python class
sp.RobustScaler()
Constructor Parameters

No constructor parameters.

Returns

JSON with transformed matrix.

Algorithm

$$x' = \frac{x - \text{median}}{\text{IQR}}$$

Example
import seraplot as sp, numpy as np
X = np.random.randn(300, 4)
X[0, :] = 100
scaler = sp.RobustScaler()
Xt = scaler.fit_transform(X)
print(Xt.mean(0))

Référence API

Nom de fonction JSON

ml_robust_scaler — alias : robust_scaler

Classe Python
sp.RobustScaler()
Paramètres du constructeur

Aucun paramètre de constructeur.

Retourne

JSON avec matrice transformed.

Algorithme

$$x' = \frac{x - \text{médiane}}{\text{IQR}}$$

Exemple
import seraplot as sp, numpy as np
X = np.random.randn(300, 4)
X[0, :] = 100
scaler = sp.RobustScaler()
Xt = scaler.fit_transform(X)
print(Xt.mean(0))

FitTransform

Transformer sklearn-compatible 🔧 Preprocessing

Unified fit_transform — dispatches to StandardScaler/MinMaxScaler/RobustScaler by name. / fit_transform unifié — dispatche vers StandardScaler/MinMaxScaler/RobustScaler par nom.

⚡ Rust-native ✓ sklearn parity
Quick start — Python 
import seraplot as sp, numpy as np
import json
X = np.random.randn(100, 3) * 5
res = json.loads(sp.ml_fit_transform(json.dumps({"X_train": X.tolist(), "scaler": "minmax"})))
💡
EN — Drop-in replacement: sp.FitTransform has the same API as sklearn.
FR — Remplacement direct : même API que sklearn, changez l'import.

API Reference

JSON function name

ml_fit_transform — aliases: fit_transform, preprocess

Python class
sp.FitTransform(scaler=standard)
Constructor Parameters
ParameterTypeDefaultDescription
scalerstrstandardScaler type: `standard`, `minmax`, `robust`.
Returns

JSON with transformed matrix.

Example
import seraplot as sp, numpy as np
import json
X = np.random.randn(100, 3) * 5
res = json.loads(sp.ml_fit_transform(json.dumps({"X_train": X.tolist(), "scaler": "minmax"})))

Référence API

Nom de fonction JSON

ml_fit_transform — alias : fit_transform, preprocess

Classe Python
sp.FitTransform(scaler=standard)
Paramètres du constructeur
ParamètreTypeDéfautDescription
scalerstrstandardType de scaler : `standard`, `minmax`, `robust`.
Retourne

JSON avec matrice transformed.

Exemple
import seraplot as sp, numpy as np
import json
X = np.random.randn(100, 3) * 5
res = json.loads(sp.ml_fit_transform(json.dumps({"X_train": X.tolist(), "scaler": "minmax"})))