Pulse Documentation

name: op-feat-poly
description: Per-row polynomial expansion x^2..x^Degree of a numeric field — the basis for a polynomial regression.
kind: operator
category: FEAT
operator: FEAT_POLY
type: reference
applies_to: process, compose, predict
examples_tags: [feature-engineering, polynomial, pre-filter, streaming-friendly]

Feature operators emit derived columns; no Response.Components.

Use when

Adds columns holding the value squared, cubed and so on up to a chosen power, so a straight-line model can bend into a curve.

Questions it answers:

• Does satisfaction rise with tenure and then level off?
• Is the effect of temperature on yield curved rather than straight?

Use something else:

• REG_OLS when you want to fit the curve itself, not just build its columns.

Params

degree — int, required. >= 2 AND <= 10 (MaxPolyDegree). Degree 1 is the original column; reference it directly downstream.

Inputs

Field — numeric u4/u8/u16/u32/u64, f32/f64, date; decimal128 via f64. No categorical, packed_bool, datetime or set_* (predict refuses them; the runtime does not yet).

Output

Degree - 1 columns <prefix>_<k> for k = 2..Degree, prefix default <field>_poly (override via Label). Each holds x^k by iterative multiplication (power *= v). The ORIGINAL column is untouched, so a linear regression over {x, x_poly_2, x_poly_3, …} is the polynomial-regression basis.

Reading the output

• value.*: One column per power k from 2 up to the degree, named <prefix>_<k> (prefix is the label, by default <field>_poly), each holding the value raised to that power: x_poly_2 is x squared, x_poly_3 is x cubed. There is no power-1 column; the original field is it.
  • Caveat: A single power column means little alone: the curve shows in a model’s coefficients on x and its powers together, and those change when the field is centred first.
  • Caveat: Even powers lose the sign: x = -3 and x = 3 both give 9 in the squared column.

Gotchas

• degree outside [2, 10] → PROCESSING_CONFIG.
• OVERFLOW WITHOUT STANDARDISATION: x^10 at |x| = 100 is already 1e20. Centre / standardise first; an orthogonal basis is out of scope for v1.
• Null inputs → null on every emitted column. Streamable per-row.

See

• pulse_examples_search tags=[polynomial], tags=[feature-engineering]
• Skills: feature-engineering, regression-modeling, op-feat-log (skew alternative), op-feat-sqrt