JuniorTheoryCommon

What does this program print?

Python

import numpy as np
from sklearn.preprocessing import StandardScaler

row = np.array([[5.1, 3.5, 1.4, 0.2]])
scaled = StandardScaler().fit_transform(row)
print(scaled.sum())

0.0

Expecting nan from dividing by a zero spread: scikit-learn does not divide by a zero standard deviation, so each value minus its own mean stays 0.
Expecting the row to pass through unchanged: fit_transform first learns the means from this one row, which equal the row itself, so every value becomes 0.

Why does a service that fits a new scaler on each request answer the same for every request?
Which object should the service apply to a raw request row instead?

Back to questions Open topic in roadmap →Answer it yourself — Quiz

Back to questions

JuniorTheoryCommon

What does this program print?

Python

import numpy as np
from sklearn.preprocessing import StandardScaler

row = np.array([[5.1, 3.5, 1.4, 0.2]])
scaled = StandardScaler().fit_transform(row)
print(scaled.sum())

0.0

Expecting nan from dividing by a zero spread: scikit-learn does not divide by a zero standard deviation, so each value minus its own mean stays 0.
Expecting the row to pass through unchanged: fit_transform first learns the means from this one row, which equal the row itself, so every value becomes 0.

Why does a service that fits a new scaler on each request answer the same for every request?
Which object should the service apply to a raw request row instead?

Back to questions Open topic in roadmap →Answer it yourself — Quiz