Ablation.drop_all()

Add a variant that removes the rows failing any step.

Usage

Source

Ablation.drop_all(steps=None)

This variant answers the question ‘does the validation as a whole help?’ Without rows_distinct() steps, its training table is the one that get_sundered_data() gives: the rows that passed every step.

Parameters

steps: list[int] | None = None

The step numbers to combine. By default, every row-level step that the ablation can use is combined.

Returns

Ablation

The Ablation object, with the variant added.

Examples

Here’s a table of temperature readings with out-of-range and missing values, validated with a step for each problem.

import pointblank as pb
import polars as pl

temps = [20 + (k * 37 % 11 - 5) * 0.6 for k in range(300)]
for k in range(0, 300, 25):
    temps[k] *= 10
for k in range(7, 300, 40):
    temps[k] = None
readings = pl.DataFrame({"temp": temps})

def evaluate(train, *, seed, variant):
    sample = train.drop_nulls().sample(fraction=1.0, with_replacement=True, seed=seed)
    return abs(sample["temp"].mean() - 20)

validation = (
    pb.Validate(data=readings)
    .col_vals_not_null(columns="temp")
    .col_vals_between(columns="temp", left=-10, right=45, na_pass=True)
    .interrogate()
)

ablation = pb.Ablation(
    validation, evaluate=evaluate, metric="error", higher_is_better=False
).drop_all()

ablation.plan()

Ablation plan
300 rows in the training table. Each version is evaluated 5 times, so the run will make 15 evaluations.
Version	Removed
Baseline
Full table Nothing removed	0 0%
Variants
All steps 1: Expect that all values in temp should not be Null. 2: Expect that values in temp should be between -10 and 45.	20 6.7%
Controls
20 random rows Rows removed at random, drawn anew for each repeat	20 6.7%

The one variant removes the 20 rows that failed either step.