Skip to content

Test from latest session #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
zingleai wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Test from latest session #4

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
53 changes: 53 additions & 0 deletions dbt/models/marts/core/schema.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,9 +83,62 @@ models:
columns:
- name: product_id
description: Unique product identifier — primary key
data_type: varchar
tests:
- unique
- not_null
- name: product_name
description: Human-readable name of the product
data_type: varchar
- name: product_category
description: Category the product belongs to (e.g. Electronics, Clothing)
data_type: varchar
- name: unit_price
description: Listed selling price per unit of the product
data_type: numeric
- name: supply_cost
description: Per-unit cost of goods from stg_supply_costs, used to compute margin
data_type: numeric
- name: unit_margin
description: Absolute margin per unit — unit_price minus supply_cost
data_type: numeric
- name: margin_pct
description: Gross margin percentage — unit_margin / unit_price × 100
data_type: numeric
- name: review_count
description: Total number of customer reviews submitted for the product
data_type: integer
- name: avg_rating
description: Average customer review rating on a 1–5 scale, derived from int_product_reviews_agg
data_type: numeric
- name: positive_reviews
description: Count of reviews with a rating >= 4 (positive sentiment threshold)
data_type: integer
- name: negative_reviews
description: Count of reviews with a rating <= 2 (negative sentiment threshold)
data_type: integer
- name: rating_tier
description: >
Product quality tier derived from avg_rating using the following scale:
'Excellent' — avg_rating >= 4.5 (top-tier, highest customer satisfaction);
'Good' — avg_rating >= 3.5 and < 4.5 (above-average satisfaction);
'Average' — avg_rating >= 2.5 and < 3.5 (neutral satisfaction);
'Poor' — avg_rating < 2.5 (below-average, needs attention).
Used for cohort segmentation in downstream models such as
marketing_excellent_rating_revenue.
data_type: varchar
- name: total_units_sold
description: Cumulative number of units sold across all orders
data_type: integer
- name: product_total_revenue
description: Total gross revenue generated by the product (units_sold × unit_price)
data_type: numeric
- name: product_gross_profit
description: Total gross profit — product_total_revenue minus total supply cost
data_type: numeric
- name: created_at
description: Timestamp when the product record was first created in the source system
data_type: timestamp

# ────────────────────────────────────────────
# dim_dates
Expand Down
42 changes: 42 additions & 0 deletions dbt/models/marts/marketing/marketing_excellent_rating_revenue.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,42 @@
-- Revenue breakdown for products in the Excellent rating cohort (avg_rating >= 4.5)
-- Filters dim_products to rating_tier = 'Excellent' and surfaces per-product
-- revenue alongside cohort-level totals and each product's percentage share.
with excellent_products as (
select
product_id,
product_name,
product_category,
unit_price,
avg_rating,
rating_tier,
review_count,
total_units_sold,
product_total_revenue,
product_gross_profit,
margin_pct
from {{ ref('dim_products') }}
where rating_tier = 'Excellent'
),

final as (
select
product_id,
product_name,
product_category,
unit_price,
avg_rating,
rating_tier,
review_count,
total_units_sold,
product_total_revenue,
product_gross_profit,
margin_pct,
sum(product_total_revenue) over () as cohort_total_revenue,
round(
product_total_revenue / nullif(sum(product_total_revenue) over (), 0) * 100,
2
) as pct_of_cohort_revenue
from excellent_products
)

select * from final
120 changes: 120 additions & 0 deletions dbt/models/marts/marketing/schema.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -173,6 +173,126 @@ models:
sla_hours: 4
tier: 2

# ────────────────────────────────────────────
# marketing_excellent_rating_revenue
# ────────────────────────────────────────────
- name: marketing_excellent_rating_revenue
description: >
Revenue analysis for products in the Excellent rating cohort (avg_rating >= 4.5).
Filters dim_products to rating_tier = 'Excellent' and surfaces per-product revenue
alongside cohort-level totals and each product's percentage share of cohort revenue.
Useful for understanding the contribution of top-rated products to overall sales.
group: marketing_analytics
access: public

config:
tags: ["active", "marketing", "tier-2", "cohort"]
materialized: table

meta:
maintainer_email: sofia.gutierrez@jaffle-shop.com
listeners:
- marketing-team@jaffle-shop.com
- jordan.blake@jaffle-shop.com
business_unit: "Marketing"
model_type: "SQL"
schedule: "Daily at 07:45 AM UTC"
schedule_cron: "45 7 * * *"
status: "active"
revised_state: "RENEWED"
expiry_date: "2027-06-30"
approved: true
approved_by: sofia.gutierrez@jaffle-shop.com
approved_date: "2026-02-23"
observe_in_airflow: true
sla_hours: 4
tier: 2
data_portal_tests:
- name: "All rows have Excellent rating tier"
type: "accepted_values"
sql: "SELECT * FROM {{ model }} WHERE rating_tier != 'Excellent'"
expected_result: "0 rows"
description: "Every row in this model must belong to the Excellent rating cohort (avg_rating >= 4.5)"
- name: "No negative product revenue"
type: "data_quality"
sql: "SELECT * FROM {{ model }} WHERE product_total_revenue < 0"
expected_result: "0 rows"
description: "Product total revenue must be non-negative for all Excellent cohort products"
- name: "Cohort total revenue is consistent across all rows"
type: "business_logic"
sql: "SELECT * FROM {{ model }} WHERE cohort_total_revenue != (SELECT SUM(product_total_revenue) FROM {{ model }})"
expected_result: "0 rows"
description: "The cohort_total_revenue window value must equal the sum of all product_total_revenue in the model"
- name: "Percentage of cohort revenue sums to 100"
type: "business_logic"
sql: "SELECT * FROM (SELECT ABS(SUM(pct_of_cohort_revenue) - 100) AS diff FROM {{ model }}) t WHERE diff > 0.01"
expected_result: "0 rows"
description: "The sum of pct_of_cohort_revenue across all Excellent products must equal 100 (within 0.01 tolerance)"
Comment on lines +226 to +230
Copy link
Copy Markdown

@coderabbitai coderabbitai bot Feb 23, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

⚠️ Potential issue | 🟡 Minor

The 0.01 tolerance for the sum-to-100 check will produce false failures for cohorts with more than a few products.

pct_of_cohort_revenue is round(..., 2), so each row carries up to ±0.005 rounding error. With N products the cumulative deviation from 100 can reach N × 0.005. A cohort of 7 equal-share products already yields a diff of 0.03, which exceeds the 0.01 threshold and triggers the test.

A simple fix is to widen the tolerance to accommodate realistic cohort sizes. If you expect up to ~50 products, a threshold of 0.25 (50 × 0.005) would be safe, or round to more decimal places in the SQL model (e.g., round(..., 4)) and tighten accordingly.

💡 Example tolerance adjustment 
-          sql: "SELECT * FROM (SELECT ABS(SUM(pct_of_cohort_revenue) - 100) AS diff FROM {{ model }}) t WHERE diff > 0.01"
+          sql: "SELECT * FROM (SELECT ABS(SUM(pct_of_cohort_revenue) - 100) AS diff FROM {{ model }}) t WHERE diff > 0.25"
-          description: "The sum of pct_of_cohort_revenue across all Excellent products must equal 100 (within 0.01 tolerance)"
+          description: "The sum of pct_of_cohort_revenue across all Excellent products must equal 100 (within 0.25 tolerance, accounting for per-row rounding to 2 decimal places)"
🤖 Prompt for AI Agents 
Verify each finding against the current code and only fix it if needed.

In `@dbt/models/marts/marketing/schema.yml` around lines 226 - 230, The
business_logic test "Percentage of cohort revenue sums to 100" is too strict
(diff > 0.01) given pct_of_cohort_revenue is rounded to 2 decimals and per-row
rounding error accumulates; update the test SQL to use a wider tolerance (e.g.,
change the WHERE clause threshold from 0.01 to 0.25) or alternatively change the
model that produces pct_of_cohort_revenue to round(..., 4) and keep a tighter
threshold; modify the SQL in the test (the sql field referencing {{ model }} and
pct_of_cohort_revenue) to implement one of these fixes.

- name: "No products with zero units sold"
type: "data_quality"
sql: "SELECT * FROM {{ model }} WHERE total_units_sold IS NULL OR total_units_sold = 0"
expected_result: "0 rows"
description: "All products in the Excellent cohort must have sold at least one unit to be included in revenue analysis"

columns:
- name: product_id
description: Unique product identifier — primary key, inherited from dim_products
data_type: varchar
tests:
- unique
- not_null
- name: product_name
description: Human-readable name of the product
data_type: varchar
tests:
- not_null
- name: product_category
description: Category the product belongs to (e.g. Electronics, Clothing)
data_type: varchar
- name: unit_price
description: Listed selling price per unit of the product
data_type: numeric
- name: avg_rating
description: Average customer review rating for the product; always >= 4.5 in this model
data_type: numeric
tests:
- not_null
- name: rating_tier
description: >
Product quality tier derived from avg_rating. Always 'Excellent' in this model
(avg_rating >= 4.5). The full tier scale in dim_products is:
Excellent (>=4.5), Good (>=3.5), Average (>=2.5), Poor (<2.5).
data_type: varchar
tests:
- not_null
- name: review_count
description: Total number of customer reviews submitted for the product
data_type: integer
- name: total_units_sold
description: Cumulative number of units sold across all orders
data_type: integer
- name: product_total_revenue
description: Total gross revenue generated by this product (units_sold × unit_price)
data_type: numeric
- name: product_gross_profit
description: Total gross profit for the product (revenue minus supply cost)
data_type: numeric
- name: margin_pct
description: Gross margin percentage — product_gross_profit / product_total_revenue × 100
data_type: numeric
Comment on lines +280 to +282
Copy link
Copy Markdown

@coderabbitai coderabbitai bot Feb 23, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

⚠️ Potential issue | 🟡 Minor

margin_pct description diverges from the authoritative dim_products definition.

The core schema defines margin_pct as "unit_margin / unit_price × 100" (Line 106-107 of core/schema.yml), while this file describes it as "product_gross_profit / product_total_revenue × 100". The two are mathematically equivalent for a constant unit price, but the inconsistent wording across files will confuse downstream consumers trying to verify the formula against the source model.

Align the description here with the dim_products wording to keep the column's meaning unambiguous.

🤖 Prompt for AI Agents 
Verify each finding against the current code and only fix it if needed.

In `@dbt/models/marts/marketing/schema.yml` around lines 280 - 282, Update the
margin_pct column description to match the canonical wording from dim_products:
replace the current text ("Gross margin percentage — product_gross_profit /
product_total_revenue × 100") with the dim_products phrasing ("unit_margin /
unit_price × 100") so the schema entry for margin_pct aligns exactly with the
authoritative definition used in dim_products; locate the margin_pct field in
this file and update its description string accordingly.

- name: cohort_total_revenue
description: >
Sum of product_total_revenue across ALL products in the Excellent rating cohort.
Computed as a window function over the full result set; the value is identical
on every row and represents the cohort's aggregate revenue contribution.
data_type: numeric
- name: pct_of_cohort_revenue
description: >
This product's share of the Excellent cohort's total revenue, expressed as a
percentage (0–100). Calculated as product_total_revenue / cohort_total_revenue × 100,
rounded to 2 decimal places. All values sum to 100 across the cohort.
data_type: numeric

# ────────────────────────────────────────────
# marketing_channel_attribution
# ────────────────────────────────────────────
Expand Down