Zero Rate

YAML type: zero_rate | Category: Data quality

Flags sustained windows where the share of near-zero days exceeds a threshold.

Description

Zero Rate looks at the proportion of near-zero days in a rolling lookback, not a single zero. Use it for sparse or zero-heavy regimes.

How it works

Zero Rate illustration

Zero Rate computes the share of near-zero days inside a rolling lookbackDays window and compares it to rateThreshold. With mode: above it alerts when zeros become too common; below alerts when zeros disappear unexpectedly.

The condition must hold for at least minDurationDays, so a single quiet day does not fire. Prefer Flatline when the series is fully stuck at a constant value.

Setup

Set lookbackDays, rateThreshold, and mode (above when too many zeros, below when zeros disappear unexpectedly).

Configure detectors on the data view Analysis tab or in YAML. Validate with Data Preview before enabling production schedules.

Settings

Parameter	Default	Description
`lookbackDays`	`30`	Window for computing zero share
`nearZeroThreshold`	`0.001`	Absolute near-zero cutoff
`rateThreshold`	`0.3`	Zero-share threshold (0-1)
`minDurationDays`	`3`	Minimum duration of the condition
`mode`	`above`	Direction -- see Allowed values

Allowed values

`mode`

Value	Meaning
`above`	Flag increases / values above the comparison (or high side of a band)
`below`	Flag decreases / values below the comparison (or low side of a band)
`both`	Flag either direction

above: near-zero rate is at least rateThreshold
below: near-zero rate is at most rateThreshold

Shared series filters

Most detectors also accept optional series filters in params:

Parameter	Purpose
`maxGapAmount`	Skip series when gap amount is too high (`null` = no limit). Catalog create-default is often `5`.
`minValue`	Skip series whose average value is below this (`null` = no limit). Catalog create-default is often `1000`.
`maxValue`	Skip series whose average value is above this (optional; omit/`null` = no limit)
`useSpecialDates`	When `true`, suppress anomalies that overlap special dates

When a filter skips a series, Data Preview shows the reason.

Configuration example

analysis:
  detectors:
    - type: zero_rate
      params:
        lookbackDays: 30
        nearZeroThreshold: 0.001
        rateThreshold: 0.3
        minDurationDays: 3
        mode: above

Using the detector

Useful when occasional zeros are normal but a high zero rate indicates failure.

Use cases

Intermittent feed dropouts
Catalog items that stop selling for long stretches
Detecting recovery when zeros suddenly vanish (mode: below)

Specifics

Prefer Flatline for long fully-stuck constant runs.