To test the consistency of a metric like

url_count

over time, you should use anomaly detection tests that monitor the value of

url_count

for unexpected changes or deviations from historical patterns. The specific monitor you choose depends on what aspect of consistency you want to track: - If

url_count

is a numeric column and you want to monitor its total value over time (e.g., sum of URLs per day), the

sum

monitor is most appropriate. This will alert you if the total

url_count

for a time bucket (like a day) is anomalous compared to previous periods. Note that

sum

is an opt-in monitor and must be explicitly configured in your test settings [all_columns_anomalies documentation](https://docs.elementary-data.com/data-tests/anomaly-detection-tests/all-columns-anomalies), [column_anomalies documentation](https://docs.elementary-data.com/data-tests/anomaly-detection-tests/column-anomalies). - If you are interested in other aspects, such as minimum, maximum, or variance of

url_count

values, you can also configure monitors like

min

,

max

, or

variance

. These will track the lowest, highest, or variability of

url_count

over time, respectively. For most use cases where you want to ensure the overall count is consistent,

sum

is the recommended monitor. Here’s an example configuration for a column anomaly test:

yaml models:   - name: your_model     columns:       - name: url_count         tests:           - elementary.column_anomalies:               column_anomalies:                 - sum

This will monitor the sum of

url_count

over your defined time buckets and alert you to anomalies in its consistency [column_anomalies configuration](https://docs.elementary-data.com/data-tests/anomaly-detection-configuration/column-anomalies). If you want to monitor all columns in a table, you can use

all_columns_anomalies

and specify

sum

as the metric for numeric columns. Summary: - Use the

sum

monitor to test the consistency of the total

url_count

over time. - Optionally, add

min

,

max

, or

variance

if you want to track other statistical properties. - Configure these in either

column_anomalies

or

all_columns_anomalies

tests, depending on your needs. For more details, see the documentation on [column_anomalies](https://docs.elementary-data.com/data-tests/anomaly-detection-tests/column-anomalies) and [all_columns_anomalies](https://docs.elementary-data.com/data-tests/anomaly-detection-tests/all-columns-anomalies).