Add_md_snapshot

Makefile

@@ -39,3 +39,11 @@ docker-run-evidence:
 		--env MDS_ENABLE_EXPORT=true \
 		--env ENVIRONMENT=docker \
 		mdsbox make run serve
+
+DATES = $(shell python -c 'from datetime import datetime, timedelta; start_date = datetime(2023, 10, 24); end_date = datetime(2024, 4, 12); date_list = [start_date + timedelta(days=x) for x in range((end_date - start_date).days + 1)]; print(" ".join(date.strftime("%Y-%m-%d") for date in date_list))')
+
+dbt-run-backfill:
+	@for date in $(DATES); do \
+		echo "Running dbt build for $$date"; \
+		(cd transform && dbt build -s tag:nba --vars "nba_start_date: $$date"); \
+	done

evidence/package.json

@@ -14,10 +14,10 @@
   },
   "type": "module",
   "dependencies": {
-    "@evidence-dev/core-components": "^4.2.0",
+    "@evidence-dev/core-components": "^4.0.1",
     "@evidence-dev/csv": "^1.0.7",
     "@evidence-dev/duckdb": "^1.0.9",
-    "@evidence-dev/evidence": "^35.0.0"
+    "@evidence-dev/evidence": "^33.0.1"
   },
   "overrides": {
     "jsonwebtoken": "9.0.0",

transform/dbt_project.yml

@@ -62,6 +62,7 @@ vars:
   nba_elo_offset: 100          # nba offset to ELO to get to ~12% home advantage
   nfl_elo_offset: 52          # nfl offset to ELO to get to 7.5% home advantage
   ncaaf_elo_offset: 52        # ncaaf offset to ELO to get to 7.5% home advantage
+  nba_start_date: '2024-04-15'    # set as of date for start of season
 
 # external files are only registered as DuckDB views when they are created, not when they are referenced.
 on-run-start:

transform/models/nba/analysis/reg_season_summary.sql

@@ -1,4 +1,8 @@
-{{ config(materialized="table") }}
+{{
+    config(
+        materialized="table"
+    )
+}}
 
 with
     cte_summary as (
@@ -41,6 +45,6 @@ select
     c.seed_5th::int || ' to ' || c.seed_95th::int as seed_range,
     c.made_postseason,
     c.made_play_in,
-    {{ var("sim_start_game_id") }} as sim_start_game_id
+    '{{ var("nba_start_date") }}'::date as nba_sim_start_date
 from cte_summary c
 left join {{ ref("nba_reg_season_actuals") }} a on a.team = c.team

transform/models/nba/analysis/season_summary.sql

@@ -1,4 +1,9 @@
-{{ config(materialized="table") }}
+{{
+    config(
+        materialized="table",
+        post_hook="COPY {{ this }} TO '../data/data_catalog/{{ this.identifier }}_{{ var('nba_start_date') }}.parquet'",
+    )
+}}
 
 select
     round(ratings.elo_rating, 0)::int

transform/models/nba/raw/nba_raw_results.sql

@@ -1,7 +1,6 @@
 with
     cte_base as (select * from {{ source("nba_dlt", "games") }}),
     cte_seed as (select * from {{ source("nba", "nba_results") }})
-
 select
     coalesce(a.date, strptime(b."Date", '%a %b %-d %Y'))::date as "date",
     b."Start (ET)" as "Start (ET)",
@@ -35,3 +34,4 @@ full outer join
     cte_seed b
     on strptime(b."Date", '%a %b %-d %Y')::date = a.date
     and b."Home/Neutral" = home.team_long
+where a.date <= '{{ var( 'nba_start_date' ) }}' 

utils/load_season_summary.py

@@ -0,0 +1,11 @@
+import os
+import duckdb
+
+# initiate the MotherDuck connection through a service token through
+con = duckdb.connect('md:?motherduck_token=' + os.environ.get('MOTHERDUCK_TOKEN'))
+
+con.sql("CREATE TABLE IF NOT EXISTS nba_history_v2.main.season_summary ( elo_rating VARCHAR(15), team VARCHAR(3), conf VARCHAR(4), record VARCHAR(10), avg_wins REAL, vegas_wins REAL, elo_vs_vegas REAL, win_range VARCHAR(15), seed_range VARCHAR(15), made_postseason INT, made_play_in INT, nba_sim_start_date DATETIME, made_playoffs INT, made_conf_semis INT, made_conf_finals INT, made_finals INT, won_finals INT, PRIMARY KEY (team,nba_sim_start_date))")
+
+con.sql("INSERT INTO nba_history_v2.main.season_summary (SELECT * FROM 'data/data_catalog/season_summar*.parquet')")
+
+print("Data Loaded to Motherduck!")