python
25 lines · 1 tab
Dr. Elena Vasquez
Apr 2026
1 tab
import pandas as pd
from sklearn.compose import ColumnTransformer
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import OneHotEncoder, StandardScaler
numeric_features = ['age', 'income', 'tenure_months']
categorical_features = ['country', 'plan_tier', 'device_type']
numeric_pipeline = Pipeline([
('imputer', SimpleImputer(strategy='median')),
('scaler', StandardScaler()),
])
categorical_pipeline = Pipeline([
('imputer', SimpleImputer(strategy='most_frequent')),
('encoder', OneHotEncoder(handle_unknown='ignore', sparse_output=False)),
])
preprocessor = ColumnTransformer([
('num', numeric_pipeline, numeric_features),
('cat', categorical_pipeline, categorical_features),
])
print(preprocessor)
1 file · python
Explain with highlit
Categoricals are where good intentions become leakage. I use one-hot encoding for low-cardinality stable fields, ordinal encoders only when order is real, and frequency or target encoders with strict cross-validation boundaries. The encoder strategy should reflect both statistical behavior and production serving constraints.
Share this code
Here's the card — post it anywhere.
