Specifying extra metrics

Now that we have computed default metrics for the current monitored tables, we can specify extra metrics to compute at table or column level. See defining your metric to set up custom metrics.

Here is an example showing how to add these metrics to the customers table.

dbt_project.yml vars

vars:  re_data:alerting_z_score: 3  re_data:schemas:    - toy_shop  re_data:monitored:    - tables:        - name: customers          time_filter: joined_at          metrics:            table: # computing table level metrics                - distinct_table_rows            column:                first_name: # computing metrics for the first_name column only                    - distinct_values                    - match_regex:                        regex: (Emily) # match rows with this name        - name: order_items          time_filter: added_at        - name: orders          time_filter: created_at      actively_monitored: true 

Metrics: distinct_table_rows, distinct_table_rows, match_regex are already defined in re_data so we just need to reference them here to let re_data now to compute them.

Once again, we run data, let's do it for the same day as previously (metrics for the first day of 2021 would be overwritten)

dbt run --models package:re_data --vars \   '{     "re_data:time_window_start": "2021-01-01 00:00:00",     "re_data:time_window_end": "2021-01-02 00:00:00"    }'

And now we can inspect ours extra metrics in SQL:

select table_name, column_name, metric, value, time_window_start, time_window_end from toy_shop_re.re_data_metricswhere metric in ('distinct_table_rows', 'distinct_values', 'match_regex');
               table_name              | column_name  |  metric             | value |  time_window_start  |   time_window_end---------------------------------------+-------------+---------------------+-------+---------------------+---------------------"postgres"."toy_shop"."customers"      | first_name   | distinct_values     |    12 | 2021-01-01 00:00:00 | 2021-01-02 00:00:00"postgres"."toy_shop"."customers"      | first_name   | match_regex         |    3  | 2021-01-01 00:00:00 | 2021-01-02 00:00:00"postgres"."toy_shop"."customers"      |              | distinct_table_rows |    15 | 2021-01-01 00:00:00 | 2021-01-02 00:00:00

What if we have a metric that is quite special for us and isn't available in re_data? No problem, we can define it ourselves and let re_data compute it. Let's check how to do it: