chore: add ibis project

002_ibis/audits/assert_positive_order_ids.sql

@@ -0,0 +1,8 @@
+AUDIT (
+  name assert_positive_order_ids,
+);
+
+SELECT *
+FROM @this_model
+WHERE
+  item_id < 0

002_ibis/config.yaml

@@ -0,0 +1,8 @@
+gateways:
+    my_gateway:
+        connection:
+            type: duckdb
+            catalogs:
+                local: 'data/local.duckdb'
+model_defaults:
+    dialect: 'duckdb'

002_ibis/constants.py

@@ -0,0 +1,3 @@
+import os
+
+DB_PATH = os.path.join(os.path.dirname(__file__), "data/local.duckdb")

002_ibis/models/full_model.sql

@@ -0,0 +1,14 @@
+MODEL (
+    name ibis.full_model,
+    kind FULL,
+    cron '@daily',
+    grain item_id,
+    audits (assert_positive_order_ids),
+);
+
+SELECT
+    item_id,
+    COUNT(DISTINCT id) AS num_orders,
+FROM
+    ibis.incremental_model
+GROUP BY item_id

002_ibis/models/ibis_full_model_python.py

@@ -0,0 +1,43 @@
+import typing as t
+from datetime import datetime
+
+import ibis  # type: ignore
+import pandas as pd
+from constants import DB_PATH  # type: ignore
+from sqlglot import exp
+
+from sqlmesh import ExecutionContext, model
+from sqlmesh.core.model import FullKind
+
+
+@model(
+    "ibis.ibis_full_model_python",
+    kind=FullKind(),
+    columns={
+        "item_id": "int",
+        "num_orders": "int",
+    },
+    audits=["assert_positive_order_ids"],
+    description="This model uses ibis to transform a `table` object and return a dataframe",
+)
+def execute(
+    context: ExecutionContext,
+    start: datetime,
+    end: datetime,
+    execution_time: datetime,
+    **kwargs: t.Any,
+) -> pd.DataFrame:
+    # get physical table name
+    upstream_model = exp.to_table(context.table("ibis.incremental_model"))
+    # connect ibis to database
+    con = ibis.duckdb.connect(DB_PATH)
+
+    # retrieve table
+    incremental_model = con.table(name=upstream_model.name, schema=upstream_model.db)
+
+    # build query
+    count = incremental_model.id.nunique()
+    aggregate = incremental_model.group_by("item_id").aggregate(num_orders=count)
+    query = aggregate.order_by("item_id")
+
+    return query.to_pandas()

002_ibis/models/ibis_full_model_sql.py

@@ -0,0 +1,28 @@
+import ibis  # type: ignore
+from ibis.expr.operations import Namespace, UnboundTable  # type: ignore
+
+from sqlmesh.core.macros import MacroEvaluator
+from sqlmesh.core.model import model
+
+
+@model(
+    "ibis.ibis_full_model_sql",
+    is_sql=True,
+    kind="FULL",
+    audits=["assert_positive_order_ids"],
+    description="This model uses ibis to generate and return a SQL string",
+)
+def entrypoint(evaluator: MacroEvaluator) -> str:
+    # create table reference
+    incremental_model = UnboundTable(
+        name="incremental_model",
+        schema={"id": "int32", "item_id": "int32", "ds": "varchar"},
+        namespace=Namespace(database="local", schema="ibis"),
+    ).to_expr()
+
+    # build query
+    count = incremental_model.id.nunique()
+    aggregate = incremental_model.group_by("item_id").aggregate(num_orders=count)
+    query = aggregate.order_by("item_id")
+
+    return ibis.to_sql(query)

002_ibis/models/incremental_model.sql

@@ -0,0 +1,18 @@
+MODEL (
+    name ibis.incremental_model,
+    kind INCREMENTAL_BY_TIME_RANGE (
+        time_column event_date
+    ),
+    start '2020-01-01',
+    cron '@daily',
+    grain (id, event_date)
+);
+
+SELECT
+    id,
+    item_id,
+    event_date,
+FROM
+    ibis.seed_model
+WHERE
+    event_date BETWEEN @start_date AND @end_date

002_ibis/models/seed_model.sql

@@ -0,0 +1,12 @@
+MODEL (
+    name ibis.seed_model,
+    kind SEED (
+        path '../seeds/seed_data.csv'
+    ),
+    columns (
+        id INTEGER,
+        item_id INTEGER,
+        event_date DATE
+    ),
+    grain (id, event_date)
+);

002_ibis/seeds/seed_data.csv

@@ -0,0 +1,8 @@
+id,item_id,event_date
+1,2,2020-01-01
+2,1,2020-01-01
+3,3,2020-01-03
+4,1,2020-01-04
+5,1,2020-01-05
+6,1,2020-01-06
+7,1,2020-01-07