Sandbox

dbt_project.yml

@@ -21,6 +21,13 @@ require-dbt-version: [">=1.0.0", "<2.0.0"]
 
 models:
   jaffle_shop:
-      materialized: table
-      staging:
+
+    materialized: table
+    staging:
+      materialized: view
+    airbnb:
+      raw:
         materialized: view
+        schema: airbnb
+
+

macros/generate_schema_name.sql

@@ -0,0 +1,14 @@
+{% macro generate_schema_name(custom_schema_name, node) -%}
+
+    {%- set default_schema = target.schema -%}
+    {%- if custom_schema_name is none -%}
+
+        {{ default_schema }}
+
+    {%- else -%}
+
+        {{ custom_schema_name | trim }}
+
+    {%- endif -%}
+
+{%- endmacro %}

macros/not_null_columns.sql

@@ -0,0 +1,7 @@
+{% macro no_nulls_in_columns(model) %}
+    SELECT * FROM {{ model }} WHERE
+    {% for col in adapter.get_columns_in_relation(model) -%}
+        {{ col.column }} IS NULL OR
+    {% endfor %}
+    FALSE
+{% endmacro %}

models/airbnb/marts/dim_listings_with_hosts.sql

@@ -0,0 +1,20 @@
+
+with dim_listings_cleansed as (
+    select * from {{ ref('stg_listings')}}
+),
+dim_hosts_cleansed  as (
+    select * from {{ ref('stg_hosts')}}
+)
+SELECT l.listing_id, 
+l.listing_name, 
+l.room_type, 
+l.minimum_nights, 
+l.price, 
+h.host_id, 
+h.created_at, 
+h.host_name, 
+h.is_superhost 
+from dim_listings_cleansed l
+ LEFT JOIN
+ dim_hosts_cleansed   h
+  on l.host_id =h.host_id

models/airbnb/marts/fct_reviews.sql

@@ -0,0 +1,16 @@
+{{
+  config(
+    materialized = 'incremental',
+    on_schema_change='fail'
+    )
+}}
+WITH stg_reviews AS (
+  SELECT * FROM {{ ref('stg_reviews') }}
+)
+SELECT 
+  *
+FROM stg_reviews
+WHERE review_text is not null
+{% if is_incremental() %}
+  AND review_date > (select max(review_date) from {{ this }})
+{% endif %}

models/airbnb/raw/src_airbnb.yml

@@ -0,0 +1,10 @@
+version: 2
+
+sources:
+  - name: airbnb
+    database: user_pchauhan 
+    schema: airbnb
+    tables:
+      - name: raw_hosts
+      - name: raw_listings
+      - name: raw_reviews 

models/airbnb/raw/src_hosts.sql

@@ -0,0 +1,24 @@
+with source as (
+
+    {#-
+    Normally we would select from the table here, but we are using seeds to load
+    our data in this project
+    #}
+    select * from {{ source('airbnb','raw_hosts') }}
+
+),
+
+renamed as (
+
+    select 
+    id as host_id,
+    name as host_name,
+    is_superhost,
+    created_at,
+    updated_at
+
+    from source
+
+)
+
+select * from renamed

models/airbnb/raw/src_listings.sql

@@ -0,0 +1,26 @@
+with source as (
+
+    {#-
+    Normally we would select from the table here, but we are using seeds to load
+    our data in this project
+    #}
+    select * from {{ source('airbnb','raw_listings') }}
+
+),
+
+renamed as (
+
+    select 
+    id as listing_id,
+    listing_url,
+    name as listing_name,
+    room_type,
+    minimum_nights,
+    host_id,
+    price as price_str,
+    CREATED_AT,
+    updated_at
+    from source
+)
+
+select * from renamed

models/airbnb/raw/src_reviews.sql

@@ -0,0 +1,23 @@
+with source as (
+
+    {#-
+    Normally we would select from the table here, but we are using seeds to load
+    our data in this project
+    #}
+    select * from {{ source('airbnb','raw_reviews') }}
+
+),
+
+renamed as (
+
+    select 
+    LISTING_ID,
+    DATE as review_date,
+    REVIEWER_NAME,
+    COMMENTS as review_text,
+    SENTIMENT as review_sentiment
+
+    from source
+)
+
+select * from renamed

models/staging/stg_customers.sql → models/airbnb/staging/stg_customers.sql

@@ -4,14 +4,14 @@ with source as (
     Normally we would select from the table here, but we are using seeds to load
     our data in this project
     #}
-    select * from {{ ref('raw_customers') }}
+    select * from {{ ref('customers1') }}
 
 ),
 
 renamed as (
 
     select
-        id as customer_id,
+        customer_id,
         first_name,
         last_name
 

models/airbnb/staging/stg_hosts.sql

@@ -0,0 +1,10 @@
+
+
+with cte as (
+    select * from {{ ref('src_hosts')}}
+)
+select 
+host_id,  NVL(host_name, 'Anonymous') as  host_name    ,
+IS_SUPERHOST,CREATED_AT,UPDATED_AT, 
+current_timestamp as staged_at
+from cte 

models/airbnb/staging/stg_listings.sql

@@ -0,0 +1,11 @@
+
+with cte as (
+    select * from {{ ref('src_listings')}}
+)
+select 
+listing_id, listing_name, room_type, host_id,
+case when minimum_nights=0 then 1 
+when minimum_nights >1 then minimum_nights end as minimum_nights,
+cast(replace(price_str,'$','') as decimal) as price ,
+CREATED_AT,UPDATED_AT
+,current_timestamp as staged_at from cte

models/staging/stg_orders.sql → models/airbnb/staging/stg_orders.sql

@@ -4,15 +4,15 @@ with source as (
     Normally we would select from the table here, but we are using seeds to load
     our data in this project
     #}
-    select * from {{ ref('raw_orders') }}
+    select * from user_pchauhan.jaffle_shop.orders
 
 ),
 
 renamed as (
 
     select
-        id as order_id,
-        user_id as customer_id,
+        order_id,
+        customer_id,
         order_date,
         status
 

models/staging/stg_payments.sql → models/airbnb/staging/stg_payments.sql

@@ -4,16 +4,16 @@ with source as (
     Normally we would select from the table here, but we are using seeds to load
     our data in this project
     #}
-    select * from {{ ref('raw_payments') }}
+    select * from {{ source('stripe','payment') }}
 
 ),
 
 renamed as (
 
     select
         id as payment_id,
-        order_id,
-        payment_method,
+        orderid as order_id,
+        paymentmethod as payment_method ,
 
         -- `amount` is currently stored in cents, so we convert it to dollars
         amount / 100 as amount

models/airbnb/staging/stg_reviews.sql

@@ -0,0 +1,9 @@
+
+
+
+with cte as (
+    select * from {{ ref('src_reviews')}}
+)
+SELECT listing_id, review_date, reviewer_name, review_text, 
+review_sentiment ,
+  current_timestamp as staged_at from cte

models/customers.sql

@@ -1,47 +1,38 @@
 with customers as (
 
-    select * from {{ ref('stg_customers') }}
+    select
+        customer_id,
+        first_name,
+        last_name
+
+    from {{ ref('stg_customers') }}
 
 ),
 
 orders as (
 
-    select * from {{ ref('stg_orders') }}
-
-),
-
-payments as (
+    select
+        order_id,
+        customer_id,
+        order_date,
+        status
 
-    select * from {{ ref('stg_payments') }}
+    from user_pchauhan.jaffle_shop.orders
 
 ),
 
 customer_orders as (
 
-        select
+    select
         customer_id,
 
-        min(order_date) as first_order,
-        max(order_date) as most_recent_order,
+        min(order_date) as first_order_date,
+        max(order_date) as most_recent_order_date,
         count(order_id) as number_of_orders
-    from orders
 
-    group by customer_id
-
-),
-
-customer_payments as (
-
-    select
-        orders.customer_id,
-        sum(amount) as total_amount
-
-    from payments
-
-    left join orders on
-         payments.order_id = orders.order_id
+    from orders
 
-    group by orders.customer_id
+    group by 1
 
 ),
 
@@ -51,19 +42,14 @@ final as (
         customers.customer_id,
         customers.first_name,
         customers.last_name,
-        customer_orders.first_order,
-        customer_orders.most_recent_order,
-        customer_orders.number_of_orders,
-        customer_payments.total_amount as customer_lifetime_value
+        customer_orders.first_order_date,
+        customer_orders.most_recent_order_date,
+        coalesce(customer_orders.number_of_orders, 0) as number_of_orders
 
     from customers
 
-    left join customer_orders
-        on customers.customer_id = customer_orders.customer_id
-
-    left join customer_payments
-        on  customers.customer_id = customer_payments.customer_id
+    left join customer_orders using (customer_id)
 
 )
 
-select * from final
+select * from final