final updates to v1 project metrics models (#1450)

* rename network as event_source

* add event_source to code metrics models

* bring artifact source from repo metrics into code metrics

* bug: rename network to event_source

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__active_developers.sql

@@ -1,5 +1,6 @@
 select
   events.project_id,
+  events.event_source,
   time_intervals.time_interval,
   'active_developer_count' as metric,
   COUNT(distinct events.from_artifact_id) as amount
@@ -10,4 +11,5 @@ where
   and events.bucket_day >= time_intervals.start_date
 group by
   events.project_id,
+  events.event_source,
   time_intervals.time_interval

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__bus_factor.sql

@@ -2,13 +2,15 @@ with all_contributions as (
   select
     project_id,
     from_artifact_id,
+    event_source,
     bucket_month,
     SUM(amount) as amount
   from {{ ref('int_events_monthly_to_project') }}
   where event_type = 'COMMIT_CODE'
   group by
     project_id,
     from_artifact_id,
+    event_source,
     bucket_month
 ),
 
@@ -22,6 +24,7 @@ aggregated_contributions as (
   select
     contributions.project_id,
     contributions.from_artifact_id,
+    contributions.event_source,
     time_intervals.time_interval,
     SUM(contributions.amount) as amount
   from contributions
@@ -32,27 +35,29 @@ aggregated_contributions as (
   group by
     contributions.project_id,
     contributions.from_artifact_id,
+    contributions.event_source,
     time_intervals.time_interval
 ),
 
 ranked_contributions as (
   select
     project_id,
+    event_source,
     time_interval,
     from_artifact_id,
     amount,
     RANK()
       over (
-        partition by project_id, time_interval
+        partition by project_id, event_source, time_interval
         order by amount desc
       ) as rank,
     SUM(amount)
       over (
-        partition by project_id, time_interval
+        partition by project_id, event_source, time_interval
       ) as total_project_amount,
     SUM(amount)
       over (
-        partition by project_id, time_interval
+        partition by project_id, event_source, time_interval
         order by amount desc
         rows between unbounded preceding and current row
       ) as cumulative_amount
@@ -61,6 +66,7 @@ ranked_contributions as (
 
 select
   project_id,
+  event_source,
   time_interval,
   'bus_factor' as metric,
   MAX(
@@ -74,4 +80,5 @@ from
   ranked_contributions
 group by
   project_id,
+  event_source,
   time_interval

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__commits_prs_issues.sql

@@ -1,5 +1,6 @@
 select
   events.project_id,
+  events.event_source,
   time_intervals.time_interval,
   CONCAT(LOWER(events.event_type), '_count') as metric,
   SUM(events.amount) as amount
@@ -16,5 +17,6 @@ where
   )
 group by
   events.project_id,
+  events.event_source,
   time_intervals.time_interval,
   events.event_type

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__contributors.sql

@@ -1,5 +1,6 @@
 select
   events.project_id,
+  events.event_source,
   time_intervals.time_interval,
   'contributor_count' as metric,
   COUNT(distinct events.from_artifact_id) as amount
@@ -14,4 +15,5 @@ where
   and events.bucket_day >= time_intervals.start_date
 group by
   events.project_id,
+  events.event_source,
   time_intervals.time_interval

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__fulltime_developers_avg.sql → warehouse/dbt/models/intermediate/metrics/code/int_code_metric__fulltime_developers_average.sql

@@ -3,6 +3,7 @@
 with dev_stats as (
   select
     events.project_id,
+    events.event_source,
     time_intervals.time_interval,
     events.from_artifact_id,
     TIMESTAMP_TRUNC(events.bucket_day, month) as bucket_month,
@@ -14,15 +15,17 @@ with dev_stats as (
     and events.bucket_day >= time_intervals.start_date
   group by
     events.project_id,
+    events.event_source,
     time_intervals.time_interval,
     events.from_artifact_id,
     TIMESTAMP_TRUNC(events.bucket_day, month)
 )
 
 select
   project_id,
+  event_source,
   time_interval,
-  'fulltime_developer_avg' as metric,
+  'fulltime_developer_average' as metric,
   (
     COUNT(distinct from_artifact_id)
     / COUNT(distinct bucket_month)
@@ -31,4 +34,5 @@ from dev_stats
 where amount >= {{ fulltime_dev_days }}
 group by
   project_id,
+  event_source,
   time_interval

warehouse/dbt/models/intermediate/metrics/code/int_code_metric__new_contributors.sql

@@ -1,6 +1,7 @@
 with user_stats as (
   select
     from_artifact_id,
+    event_source,
     project_id,
     min(bucket_day) as first_day
   from {{ ref('int_events_daily_to_project') }}
@@ -12,11 +13,13 @@ with user_stats as (
     )
   group by
     from_artifact_id,
+    event_source,
     project_id
 )
 
 select
   events.project_id,
+  events.event_source,
   time_intervals.time_interval,
   'new_contributor_count' as metric,
   count(
@@ -31,6 +34,7 @@ inner join user_stats
   on
     events.from_artifact_id = user_stats.from_artifact_id
     and events.project_id = user_stats.project_id
+    and events.event_source = user_stats.event_source
 cross join {{ ref('int_time_intervals') }} as time_intervals
 where
   events.event_type in (
@@ -41,4 +45,5 @@ where
   and events.bucket_day >= time_intervals.start_date
 group by
   events.project_id,
+  events.event_source,
   time_intervals.time_interval

warehouse/dbt/models/intermediate/metrics/int_code_metrics_by_project.sql

@@ -5,14 +5,16 @@ with metrics as (
   union all
   select * from {{ ref('int_code_metric__contributors') }}
   union all
-  select * from {{ ref('int_code_metric__fulltime_developers_avg') }}
+  select *
+  from {{ ref('int_code_metric__fulltime_developers_average') }}
   union all
   select * from {{ ref('int_code_metric__new_contributors') }}
 ),
 
 aggs as (
   select
     project_id,
+    event_source,
     SUM(
       case
         when
@@ -97,53 +99,81 @@ aggs as (
     SUM(
       case
         when
-          metric = 'fulltime_developer_avg'
+          metric = 'fulltime_developer_average'
           and time_interval = '6 MONTHS'
           then amount
         else 0
       end
-    ) as fulltime_developer_avg_6_months
+    ) as fulltime_developer_average_6_months
   from metrics
-  group by project_id
+  group by
+    project_id,
+    event_source
 ),
 
 repos as (
   select
     project_id,
+    artifact_source as event_source,
     MIN(first_commit_time) as first_commit_date,
     MAX(last_commit_time) as last_commit_date,
     COUNT(distinct artifact_id) as repository_count,
     SUM(star_count) as star_count,
     SUM(fork_count) as fork_count
   from {{ ref('int_repo_metrics_by_project') }}
   --WHERE r.is_fork = false
-  group by project_id
+  group by
+    project_id,
+    artifact_source
+),
+
+code_metrics as (
+  select
+    repos.*,
+    aggs.* except (project_id, event_source)
+  from repos
+  left join aggs
+    on
+      repos.project_id = aggs.project_id
+      and repos.event_source = aggs.event_source
+),
+
+project_metadata as (
+  select
+    project_id,
+    project_source,
+    project_namespace,
+    project_name,
+    display_name,
+    'GITHUB' as event_source
+  from {{ ref('int_projects') }}
+
 )
 
 select
-  int_projects.project_id,
-  int_projects.project_source,
-  int_projects.project_namespace,
-  int_projects.project_name,
-  int_projects.display_name,
-  repos.first_commit_date,
-  repos.last_commit_date,
-  repos.repository_count,
-  repos.star_count,
-  repos.fork_count,
-  aggs.contributor_count,
-  aggs.contributor_count_6_months,
-  aggs.new_contributor_count_6_months,
-  aggs.fulltime_developer_avg_6_months,
-  aggs.active_developer_count_6_months,
-  aggs.commit_count_6_months,
-  aggs.opened_pull_request_count_6_months,
-  aggs.merged_pull_request_count_6_months,
-  aggs.opened_issue_count_6_months,
-  aggs.closed_issue_count_6_months,
-  'GITHUB' as repository_source
-from {{ ref('int_projects') }}
-left join aggs
-  on int_projects.project_id = aggs.project_id
-left join repos
-  on int_projects.project_id = repos.project_id
+  project_metadata.project_id,
+  project_metadata.project_source,
+  project_metadata.project_namespace,
+  project_metadata.project_name,
+  project_metadata.display_name,
+  project_metadata.event_source,
+  code_metrics.first_commit_date,
+  code_metrics.last_commit_date,
+  code_metrics.repository_count,
+  code_metrics.star_count,
+  code_metrics.fork_count,
+  code_metrics.contributor_count,
+  code_metrics.contributor_count_6_months,
+  code_metrics.new_contributor_count_6_months,
+  code_metrics.fulltime_developer_average_6_months,
+  code_metrics.active_developer_count_6_months,
+  code_metrics.commit_count_6_months,
+  code_metrics.opened_pull_request_count_6_months,
+  code_metrics.merged_pull_request_count_6_months,
+  code_metrics.opened_issue_count_6_months,
+  code_metrics.closed_issue_count_6_months
+from project_metadata
+left join code_metrics
+  on
+    project_metadata.project_id = code_metrics.project_id
+    and project_metadata.event_source = code_metrics.event_source

warehouse/dbt/models/intermediate/metrics/int_onchain_metrics_by_project.sql

@@ -32,7 +32,7 @@ with metrics as (
 aggs as (
   select
     project_id,
-    network,
+    event_source,
     SUM(
       case
         when
@@ -162,7 +162,7 @@ aggs as (
   from metrics
   group by
     project_id,
-    network
+    event_source
 )
 
 select

warehouse/dbt/models/intermediate/metrics/int_repo_metrics_by_project.sql

@@ -46,6 +46,7 @@ select
   int_artifacts_by_project.artifact_id,
   int_artifacts_by_project.artifact_namespace,
   int_artifacts_by_project.artifact_name,
+  int_artifacts_by_project.artifact_source,
   repo_snapshot.is_fork,
   repo_snapshot.fork_count,
   repo_snapshot.star_count,

warehouse/dbt/models/intermediate/metrics/onchain/int_onchain_metric__active_addresses.sql

@@ -1,6 +1,6 @@
 select
   events.project_id,
-  events.event_source as network,
+  events.event_source,
   time_intervals.time_interval,
   'address_count' as metric,
   COUNT(distinct events.from_artifact_id) as amount

warehouse/dbt/models/intermediate/metrics/onchain/int_onchain_metric__active_contracts.sql

@@ -1,6 +1,6 @@
 select
   events.project_id,
-  events.event_source as network,
+  events.event_source,
   time_intervals.time_interval,
   'active_contract_count' as metric,
   COUNT(distinct events.to_artifact_id) as amount

warehouse/dbt/models/intermediate/metrics/onchain/int_onchain_metric__days_since_first_transaction.sql

@@ -1,6 +1,6 @@
 select
   events.project_id,
-  events.event_source as network,
+  events.event_source,
   time_intervals.time_interval,
   'days_since_first_transaction' as metric,
   MAX(

warehouse/dbt/models/intermediate/metrics/onchain/int_onchain_metric__gas_fees.sql

@@ -1,6 +1,6 @@
 select
   events.project_id,
-  events.event_source as network,
+  events.event_source,
   time_intervals.time_interval,
   'gas_fees' as metric,
   SUM(events.amount / 1e18) as amount