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

name: op-agg-rmse
description: Root mean squared error of a prediction against an actual — sqrt(mean((y - prediction)^2)).
kind: operator
category: AGG
operator: AGG_RMSE
type: reference
applies_to: process, compose, predict
examples_tags: [regression, compose]

Use when

Typical size of a prediction’s miss in the actual’s own units, with large misses counting more: the root of the average squared miss.

Questions it answers:

  • How many units off is the sales model on the held-out weeks, with big misses weighing more?
  • Which of two models misses less on the same rows?

Use something else:

  • AGG_MAE when every unit of miss should count the same, so a few large misses do not dominate.

Params

prediction_field — required; the predicted column (typically an ATTR_REG_PREDICT score column). Missing, or a non-numeric column on either side (categorical, set, date, datetime, decimal128), is PROCESSING_CONFIG; an unknown name SERVICE_VALIDATION — predict and runtime alike.

Weight-aware: √(Σw·(y − ŷ)² / Σw), the root of the pooled MSE. A decimal128 side under a weight → PULSE_WEIGHT_UNSUPPORTED.

Inputs

Field is the actual: u4–u64, f32/f64, packed_bool. No train/holdout param — split with a filter or GROUP_RANGE range grouper.

Output

Scalar float64 in the actual’s units. Rows with a null on either side are skipped; none left → NaN (null).

Reading the output

  • value: The square root of the average squared miss between prediction and actual, in the actual’s own units. 0 means every prediction is exact; it is never below AGG_MAE on the same rows.
    • Caveat: Squaring makes a few large misses dominate it; a gap well above AGG_MAE on the same rows says the misses are uneven.
    • Caveat: It has the actual’s units, so compare it only between models scored on the same rows and target.

Components

Floor {n, n_null} (on field only) (weighted adds sum_weights, n_eff, n_weight_invalid) plus loss_sum (Σw·(y − ŷ)²), weight_sum, n_scored, n_null_prediction, mse, rmse. Mergeable (pooled sums, root at finalize); streams.

Gotchas

  • Squaring lets a few large misses dominate; always ≥ AGG_MAE on the same rows.

See