Read coefficients after encoding

Understand the idea

Dropping one category gives an explicit reference for linear coefficients with an intercept. Category coefficients compare with that reference, conditional on other inputs. The fitted encoder exposes categories_ and drop_idx_. Read its category list at the dropped index to find the reference; do not guess it from frequency or original row order.

Align coefficients with encoded feature names and a reference category.Scroll the diagram horizontally if needed.

What the syntax does

model.named_steps['prepare']: named_steps maps the authored pipeline names to their fitted objects.
Reading a contained estimator’s fitted coefficients or preparation schema does not fit again.
Use the exact step names from the supplied recipe.
The prepare step contains fitted preparation; model contains the prediction estimator.
A ColumnTransformer also exposes its named_transformers_ when inspecting encoded categories.
drop='first': Omits one category indicator so coefficients compare other categories with that reference.
encoder.categories_[0]: The first categorical input’s fitted category list, in encoder order.

Supplied for this practice

These objects are prepared afresh before your editor runs.

X: Feature dataframe for the declared population, preserving row indices.
X = df[['distance', 'weight', 'service', 'weekend']]
y: Target series, aligned with X.
y = df['duration']
X_train / X_test / y_train / y_test: Training feature rows from the declared split. Final-test feature rows from the declared split. Training targets, aligned with X_train. Final-test targets, aligned with X_test.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
prepare: Already fitted on X_train, y_train as part of model. Supplied prepare; its definition below specifies the values and role. Fitted on X_train, y_train; reuse without refitting.
model: Already fitted on X_train, y_train. Fitted estimator or pipeline for the requested model family. Fitted on X_train, y_train; reuse without refitting.

Outputs you create

answer: The requested result, with the values, labels and row order described by this task.

MIX60 · first 8 rows
`distance`	`weight`	`service`	`weekend`	`duration`
15.7052	6.75035	standard	0	61.7704
9.33869	4.81674	express	1	35.1694
17.3134	5.73931	economy	0	59.0766
14.25	7.69699	standard	1	56.5785
2.78937	6.42024	express	0	19.5577
19.5368	5.62508	economy	1	77.0289
15.4617	5.68023	standard	0	56.8109
15.9352	3.17871	express	1	52.08

Your task

Identify the omitted service reference from the fitted encoder.

answer

Hints

Think: The omitted indicator establishes the category against which the encoded coefficient is compared.

Tools: named_transformers_, categories_ and drop_idx_.

Approach: Reach the fitted service encoder and use its dropped-position metadata to identify the reference.

Explained solution

encoder=model.named_steps['prepare'].named_transformers_['service']
answer=encoder.categories_[0][encoder.drop_idx_[0]]

Reading fitted metadata avoids assuming an alphabetical reference or guessing from input order.