https://linen.dev logo
Join Slack
Powered by
<@U088WCP3B4P> How can I go about replacing this t...
# ask-elementary-ai
c
@kapa.ai How can I go about replacing this test in my project?
Copy code
{% test volume_anomalies(model, timestamp_column, where_expression, anomaly_sensitivity, anomaly_direction, min_training_set_size, time_bucket, days_back, backfill_days, seasonality, sensitivity, ignore_small_changes, fail_on_zero, detection_delay, anomaly_exclude_metrics, detection_period, training_period) %}
  {{ config(tags = ['elementary-tests']) }}

  {{ elementary.test_table_anomalies(
      model=model,
      table_anomalies=["row_count"],
      freshness_column=none,
      timestamp_column=timestamp_column,
      where_expression=where_expression,
      anomaly_sensitivity=anomaly_sensitivity,
      anomaly_direction=anomaly_direction,
      min_training_set_size=min_training_set_size,
      time_bucket=time_bucket,
      days_back=days_back,
      backfill_days=backfill_days,
      seasonality=seasonality,
      sensitivity=sensitivity,
      ignore_small_changes=ignore_small_changes,
      fail_on_zero=fail_on_zero,
      detection_delay=detection_delay,
      anomaly_exclude_metrics=anomaly_exclude_metrics,
      detection_period=detection_period,
      training_period=training_period
    )
  }}
{% endtest %}
k
A message from kapa.ai
To replace your custom 
volume_anomalies
test that wraps 
elementary.test_table_anomalies
with the modern, recommended approach, you should use the built-in 
elementary.volume_anomalies
test in your 
schema.yml
or model configuration. The 
elementary.volume_anomalies
test is now the preferred and documented way to monitor row count anomalies, and it supports all the relevant configuration parameters directly. Here’s how you can translate your custom test to the new format:
Copy code
yaml models:   - name: <model_name>     tests:       - elementary.volume_anomalies:           timestamp_column: <timestamp_column>           where_expression: <where_expression>           anomaly_sensitivity: <anomaly_sensitivity>           anomaly_direction: <anomaly_direction>           min_training_set_size: <min_training_set_size>           time_bucket:             period: <period>             count: <count>           days_back: <days_back>           backfill_days: <backfill_days>           seasonality: <seasonality>           sensitivity: <sensitivity>           ignore_small_changes:             spike_failure_percent_threshold: <value>             drop_failure_percent_threshold: <value>           fail_on_zero: <fail_on_zero>           detection_delay:             period: <period>             count: <count>           anomaly_exclude_metrics: <anomaly_exclude_metrics>           detection_period:             period: <period>             count: <count>           training_period:             period: <period>             count: <count>           tags: ['elementary-tests']
Key points: - All the parameters you used in your custom test are supported as arguments to 
elementary.volume_anomalies
. - You no longer need to use a custom Jinja test or call 
elementary.test_table_anomalies
directly. - The 
elementary.volume_anomalies
test is the direct replacement for the 
table_anomalies
test with 
row_count
and is the recommended approach going forward. Backward compatibility for the old approach exists, but it is no longer documented or recommended for new projects [Test configuration](https://docs.elementary-data.com/data-tests/anomaly-detection-tests/volume-anomalies#test-configuration) [Slack clarification](https://slack.com/archives/C02CTC89LAX/p1692128245054759). If you have any parameters that are not directly supported, consult the [anomaly detection configuration documentation](https://docs.elementary-data.com/data-tests/anomaly-detection-configuration/anomaly-params) for the full list of supported options. Summary: Replace your custom test with a YAML-based 
elementary.volume_anomalies
test in your model or schema file, mapping your parameters accordingly. This is the modern, supported, and documented approach.
32 Views
Open in Slack
PreviousNext
Powered by