← All workshopsRead Saving the Whole Prediction Procedure

Deploying and Maintaining Models

Reproduce the saved predictor

BOTIE

Transform the raw feature vector using the retained means and scales without refitting either.
Ready to test
0 passed · 0 failed · 2 not run

An empty floor. Start with the information your machine needs.

Received result

Your answer will arrive here.

Examples, experiments and workspace controls
Live values

Connect a tool's inputs to see what it produces. You can try a branch before connecting it to Result.

Try your own inputs

Use a copy of the example to try your own data. Keep the problem's assumptions in mind, and predict the effect of a change before checking the output. You can also add another branch to compare two ways of doing the calculation.

Start at an output arrow, then click or drag to an input. Drag empty space to pan. Delete removes a selection. Ctrl/Cmd+Z undoes a change.
Context for this challenge

Transform the raw feature vector using the retained means and scales without refitting either.

Botie has raw feature values and every fitted quantity used during training.

Inspect the component calculation
Multiply(
  Value: Subtract(
    Starting value: Raw feature vector
    Value to remove: Training means
  )
  Multiplier: Take positive reciprocals(
    Positive values: Positive training scales
  )
)
Examples and results
Raw feature vector
  1. 012
  2. 125
Training means
  1. 010
  2. 120
Positive training scales
  1. 02
  2. 15
Required result
  1. 01
  2. 11
Raw feature vector
  1. 08
  2. 115
Training means
  1. 010
  2. 120
Positive training scales
  1. 02
  2. 15
Required result
  1. 0-1
  2. 1-1
Try a fresh construction

This clears this draft, including its undo history. Earlier successful attempts stay recorded. Build again without opening hints to record an independent attempt.

Stuck? Inspect a worked construction

The tools below form one possible solution. You can assemble it, follow its signals, and undo that change to return to your own construction.

Assumptions and test precision

Feature order and units must match the saved contract. This example is a linear predictor, not arbitrary model serialisation. Numerical tests allow a difference of 0.001. Lists and structures must also have the right shape.

Your last tested pieces can start the full construction. An existing draft there keeps its own pieces.

Production data still uses the training scale

This check uses the pieces from “Replay preparation and prediction from saved quantities”.

BOTIE

Botie receives three raw features for one new observation. Apply the stored training means and scales before the fitted weights and intercept. Do not calculate new preprocessing statistics from this one production row.
Raw feature vector
  1. 014
  2. 115
  3. 28
Training means
  1. 010
  2. 120
  3. 28
Positive training scales
  1. 02
  2. 15
  3. 24
Saved weights
  1. 01.5
  2. 1-2
  3. 26
Saved bias
-1

Use a number, text, or an array matching the requested result. For a matrix, put each row in brackets. Saving your prediction runs your current construction on these inputs and checks the earlier examples too.

Build a construction in the workspace above first. This case will test those pieces.

What would change if a request refitted its own mean and scale?

Save feature order, units, preprocessing state, and model parameters as one reproducible procedure.

Feature order and units must match the saved contract. This example is a linear predictor, not arbitrary model serialisation.

Work through the Python challenges →
Continue the courseChecking What Arrives at the Model →
← Reweighting Records Without Inventing Missing Evidence