updated logic to include data cleaning for datetime

This commit is contained in:
CIEF ACC1
2023-08-23 09:02:33 +00:00
parent c5aff88531
commit f32ccded38
@@ -1,8 +1,15 @@
-- IMPORTS
-- VARIABLES
{% set order_status = dbt_utils.get_column_values(
table=ref('stg_exchange__transaction_order_logs'),
column='status') %}
column='status')
%}
{% set booking_status = dbt_utils.get_column_values(
table=ref('stg_exchange__booking_logs'),
column='status')
%}
-- IMPORTS
WITH transaction_orders AS (
SELECT * FROM {{ ref('int_exchange__transaction_orders_get_user_ids') }}
),
@@ -11,32 +18,208 @@ transaction_order_logs AS (
SELECT * FROM {{ ref('stg_exchange__transaction_order_logs') }}
),
-- LOGICS
transaction_costs AS (
SELECT * FROM {{ ref('stg_exchange__transaction_costs') }}
),
transaction_cost_logs AS (
SELECT * FROM {{ ref('stg_exchange__transaction_cost_logs') }}
),
bookings AS (
SELECT * FROM {{ ref('int_exchange__bookings_get_estimate_booking_values') }}
),
booking_logs AS (
SELECT * FROM {{ ref('stg_exchange__booking_logs') }}
),
-- LOGIC
order_logs_join_bookings AS (
SELECT
transaction_order_logs.*,
bookings.service_type,
ROW_NUMBER() OVER (PARTITION BY transaction_order_id
ORDER BY transaction_order_logs.updated_datetime DESC,
transaction_order_logs.transaction_order_id DESC,
transaction_order_logs.transaction_order_log_id DESC) AS rank_index
FROM transaction_order_logs
LEFT JOIN bookings
ON (transaction_order_logs.booking_id = bookings.booking_id)
),
transaction_order_order_status_datetime AS (
SELECT
transaction_order_id,
payment_method,
service_type,
MAX(rank_index) AS status_count,
MAX(CASE WHEN rank_index = 1 THEN status END) AS final_order_status,
MIN(created_datetime) AS order_created_datetime,
{%- for status in order_status %}
MAX(CASE WHEN status = '{{status}}' THEN updated_datetime END) AS order_{{status.lower()}}_datetime {%- if not loop.last %},{% endif -%}
{% if status != 'APPROVED' %}
MAX(CASE WHEN status = '{{status}}' THEN updated_datetime END) AS order_{{status.lower()}}_datetime
{% else %}
MIN(CASE WHEN status = '{{status}}' THEN updated_datetime END) AS order_{{status.lower()}}_datetime
{% endif %}
{%- if not loop.last %},{% endif -%}
{% endfor %}
FROM transaction_order_logs
FROM order_logs_join_bookings
GROUP BY
transaction_order_id
transaction_order_id,
payment_method,
service_type
),
transaction_order_datetime_imputation AS (
SELECT
transaction_order_id,
payment_method,
service_type,
final_order_status,
order_created_datetime AS new_order_created_datetime,
order_rejected_datetime AS new_order_rejected_datetime,
order_suspended_datetime AS new_order_suspended_datetime,
order_expired_datetime AS new_order_expired_datetime,
-- imputation for ORDER_PENDING_SUBMISSION DATETIME
COALESCE(order_pending_submission_datetime, order_created_datetime) AS new_order_pending_submission_datetime,
/*
Data manipulation for ORDER_PENDING_VERIFICATION_DATETIME
- To improve data consistency
*/
CASE
-- When final order status is not pending verification, return null
WHEN final_order_status NOT IN ('PENDING_VERIFICATION', 'APPROVED', 'COMPLETED') THEN null
ELSE COALESCE(order_pending_verification_datetime, new_order_pending_submission_datetime)
END AS new_order_pending_verification_datetime,
/*
Data manipulation for ORDER_APPROVED_DATETIME
- To improve data consistency
- All orders from year 2023 has order_approved datetime
*/
CASE
-- When final order status is not approved, return null
WHEN order_created_datetime >= '2023-01-01'AND final_order_status NOT IN ('APPROVED', 'COMPLETED') THEN null
ELSE COALESCE(order_approved_datetime, new_order_pending_verification_datetime)
END AS new_order_approved_datetime,
/*
Missing value imputation for order_completed_datetime
- Missing value caused by inconsistency and backlog from exchange system
- Imputed dataset used for estimation of on-time delivery
*/
CASE
-- When final order status is not completed, return null, imputation will be done in next code block
WHEN order_created_datetime >= '2023-01-01'AND final_order_status NOT IN ('COMPLETED') THEN null
WHEN order_created_datetime < '2023-01-01' THEN order_completed_datetime
/*
Imputation for order_approved_datetime for 1688 payment
- If payment_method is 'cash', add median of 53 minute to order_approved datetime
- If payment_method is 'wallet', add median of 569 minute to order_approved datetime
- If payment_method is 'cheque', reuse order_completed datetime, no orders using cheque in 1688 payment
- If payment_method is 'payment_gateway', add median of 433 minute to order_approved datetime
*/
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1688 PAYMENT' AND payment_method = 'CASH'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 53, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1688 PAYMENT' AND payment_method = 'WALLET'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 569, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1688 PAYMENT' AND payment_method = 'CHEQUE'
THEN order_completed_datetime
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1688 PAYMENT' AND payment_method = 'PAYMENT_GATEWAY'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 433, new_order_approved_datetime))
/*
Imputation for order_approved_datetime for 1 day transfer
- If payment_method is 'cash', add median of 59 minute to order_approved datetime
- If payment_method is 'wallet', add median of 206 minute to order_approved datetime
- If payment_method is 'cheque', reuse order_completed datetime, all orders have completed datetime
- If payment_method is 'payment_gateway', add median of 214.5 minute to order_approved datetime
*/
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1 DAY TRANSFER' AND payment_method = 'CASH'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 59, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1 DAY TRANSFER' AND payment_method = 'WALLET'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 206, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1 DAY TRANSFER' AND payment_method = 'CHEQUE'
THEN order_completed_datetime
WHEN order_created_datetime >= '2023-01-01' AND service_type = '1 DAY TRANSFER' AND payment_method = 'PAYMENT_GATEWAY'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 214.5, new_order_approved_datetime))
/*
Imputation for order_approved_datetime for 3 days transfer
- If payment_method is 'cash', add median of 2702 minute to order_approved datetime
- If payment_method is 'wallet', add median of 4502.92 minute to order_approved datetime
- If payment_method is 'cheque', reuse order_completed datetime, all orders have completed datetime
- If payment_method is 'payment_gateway', add median of 5253 minute to order_approved datetime
*/
WHEN order_created_datetime >= '2023-01-01' AND service_type = '3 DAYS TRANSFER' AND payment_method = 'CASH'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 2702, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '3 DAYS TRANSFER' AND payment_method = 'WALLET'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 4502.92, new_order_approved_datetime))
WHEN order_created_datetime >= '2023-01-01' AND service_type = '3 DAYS TRANSFER' AND payment_method = 'CHEQUE'
THEN order_completed_datetime
WHEN order_created_datetime >= '2023-01-01' AND service_type = '3 DAYS TRANSFER' AND payment_method = 'PAYMENT_GATEWAY'
THEN COALESCE(order_completed_datetime, DATEADD(minute, 5253, new_order_approved_datetime))
END AS new_order_completed_datetime
FROM transaction_order_order_status_datetime
),
transaction_cost_status_datetime AS (
SELECT
transaction_cost_id,
MIN(created_datetime) AS cost_created_datetime,
{%- for status in order_status %}
MAX(CASE WHEN status = '{{status}}' THEN updated_datetime END) AS cost_{{status.lower()}}_datetime {%- if not loop.last %},{% endif -%}
{% endfor %}
FROM transaction_cost_logs
GROUP BY
transaction_cost_id
),
booking_status_datetime AS (
SELECT
booking_id,
MIN(created_datetime) AS booking_created_datetime,
{%- for status in booking_status %}
MAX(CASE WHEN status = '{{status}}' THEN updated_datetime END) AS booking_{{status.lower()}}_datetime {%- if not loop.last %},{% endif -%}
{% endfor %}
FROM booking_logs
GROUP BY
booking_id
),
transaction_orders_lists AS (
SELECT
transaction_orders.transaction_order_id,
transaction_orders.booking_id,
transaction_orders.bank_id,
transaction_orders.company_id,
transaction_orders.user_id,
transaction_orders.status AS order_status,
transaction_orders.status,
transaction_orders.transaction_type,
transaction_orders.payment_method,
transaction_orders.base_currency_id,
@@ -44,7 +227,6 @@ transaction_orders_lists AS (
transaction_orders.base_to_quote_currency_exchange_rate,
transaction_orders.payment_reference,
transaction_orders.bill_number,
transaction_orders.expired_datetime AS order_expired_datetime,
transaction_orders.base_value,
transaction_orders.quote_value,
@@ -67,12 +249,13 @@ transaction_orders_lists AS (
(value_rm + service_charge_rm + tax_rm) AS total_value_rm,
transaction_orders.expired_datetime AS order_expected_expired_datetime,
transaction_orders.deleted_datetime AS order_deleted_datetime,
transaction_orders.updated_datetime AS order_updated_datetime,
{% for column_name in ["company", "user"] %}
CASE
WHEN ROW_NUMBER() OVER (PARTITION BY transaction_orders.{{column_name}}_id ORDER BY transaction_orders.created_datetime) = 1 THEN 1
WHEN ROW_NUMBER() OVER (PARTITION BY transaction_orders.{{column_name}}_id ORDER BY transaction_orders.created_datetime, transaction_orders.transaction_order_id) = 1 THEN 1
ELSE 0
END AS is_first_time_{{column_name}},
@@ -87,72 +270,294 @@ transaction_orders_lists AS (
END AS is_first_time_{{column_name}}_completed,
{% endfor %}
transaction_order_order_status_datetime.order_created_datetime,
transaction_order_datetime_imputation.new_order_created_datetime AS order_created_datetime,
-- transaction_order_datetime_imputation.order_expired_datetime,
{%- for status in order_status %}
transaction_order_order_status_datetime.order_{{status.lower()}}_datetime,
transaction_order_datetime_imputation.new_order_{{status.lower()}}_datetime AS order_{{status.lower()}}_datetime,
{% endfor %}
'{{ modules.datetime.datetime.now(modules.pytz.timezone("Asia/Kuala_Lumpur")) }}' AS _dbt_ran_datetime
LEAD(transaction_order_datetime_imputation.new_order_created_datetime) OVER
(PARTITION BY company_id ORDER BY transaction_order_datetime_imputation.new_order_created_datetime,
transaction_order_datetime_imputation.transaction_order_id) AS next_order_created_datetime
FROM transaction_orders
LEFT JOIN transaction_order_order_status_datetime
ON (transaction_orders.transaction_order_id = transaction_order_order_status_datetime.transaction_order_id)
LEFT JOIN transaction_order_datetime_imputation
ON (transaction_orders.transaction_order_id = transaction_order_datetime_imputation.transaction_order_id)
),
transaction_costs_lists AS (
SELECT
transaction_costs.transaction_cost_id,
transaction_costs.transaction_order_id,
transaction_costs.supplier_company_id,
transaction_costs.bank_id,
transaction_costs.base_currency_id,
transaction_costs.quote_currency_id,
transaction_costs.transaction_type,
transaction_costs.payment_method,
transaction_costs.status,
transaction_costs.payment_reference,
transaction_costs.bill_number,
transaction_costs.base_value,
transaction_costs.quote_value,
transaction_costs.base_to_quote_currency_exchange_rate,
transaction_costs.base_tax,
transaction_costs.base_service_charge,
-- use for future if base currency id other from '1'
IFF(transaction_costs.base_currency_id = '1',
transaction_costs.base_value,
-9999999999) AS value_rm,
-- use for future if base currency id other from '1'
IFF(transaction_costs.base_currency_id = '1',
transaction_costs.base_tax,
-9999999999) AS tax_rm,
-- use for future if base currency id other from '1'
IFF(transaction_costs.base_currency_id = '1',
transaction_costs.base_service_charge,
-9999999999) AS service_charge_rm,
(value_rm + service_charge_rm + tax_rm) AS total_value_rm,
transaction_costs.expired_datetime,
transaction_costs.deleted_datetime,
transaction_costs.created_datetime,
transaction_costs.updated_datetime,
{%- for status in order_status %}
transaction_cost_status_datetime.cost_{{status.lower()}}_datetime,
{% endfor %}
transaction_cost_status_datetime.cost_created_datetime
FROM transaction_costs
LEFT JOIN transaction_cost_status_datetime
ON (transaction_costs.transaction_cost_id = transaction_cost_status_datetime.transaction_cost_id)
),
bookings_lists AS (
SELECT
bookings.*,
{% for column_name in ["company", "user"] %}
CASE
WHEN ROW_NUMBER() OVER (PARTITION BY bookings.{{column_name}}_id ORDER BY bookings.created_datetime, bookings.booking_id) = 1 THEN 1
ELSE 0
END AS is_first_time_{{column_name}},
CASE
WHEN bookings.status = 'COMPLETED'
AND
MIN(IFF(bookings.status = 'COMPLETED', bookings.created_datetime, NULL))
OVER (PARTITION BY bookings.{{column_name}}_id)
= bookings.created_datetime
THEN 1
ELSE 0
END AS is_first_time_{{column_name}}_completed,
{% endfor %}
booking_status_datetime.booking_created_datetime,
{%- for status in booking_status %}
booking_status_datetime.booking_{{status.lower()}}_datetime {%- if not loop.last %},{% endif -%}
{% endfor %}
FROM bookings
LEFT JOIN booking_status_datetime
ON (bookings.booking_id = booking_status_datetime.booking_id)
),
bookings_orders_costs_join AS (
SELECT
bookings_lists.booking_id,
bookings_lists.booking_marking_id,
bookings_lists.user_id,
bookings_lists.fix_currency_id,
bookings_lists.quote_currency_id,
bookings_lists.base_currency_id,
bookings_lists.service_type,
bookings_lists.status AS booking_status,
bookings_lists.is_first_time_company AS is_first_time_booking_company,
bookings_lists.is_first_time_company_completed AS is_first_time_booking_company_completed,
bookings_lists.is_first_time_user AS is_first_time_booking_user,
bookings_lists.is_first_time_user_completed AS is_first_time_booking_user_completed,
bookings_lists.estimate_base_to_quote_currency_exchange_rate AS estimate_booking_base_to_quote_currency_exchange_rate,
bookings_lists.estimate_quote_value AS estimate_booking_quote_value,
bookings_lists.estimate_base_value AS estimate_booking_base_value,
bookings_lists.estimate_value_rm AS estimate_booking_value_rm,
bookings_lists.booking_created_datetime,
bookings_lists.booking_approved_datetime,
bookings_lists.booking_completed_datetime,
bookings_lists.booking_suspended_datetime,
transaction_orders_lists.transaction_order_id AS order_id,
transaction_orders_lists.company_id,
transaction_orders_lists.bank_id,
transaction_orders_lists.transaction_type AS order_transaction_type,
transaction_orders_lists.payment_method AS order_payment_method,
transaction_orders_lists.status AS order_status,
transaction_orders_lists.payment_reference AS order_payment_reference,
transaction_orders_lists.bill_number AS order_bill_number,
transaction_orders_lists.is_first_time_company AS is_first_time_order_company,
transaction_orders_lists.is_first_time_company_completed AS is_first_time_order_company_completed,
transaction_orders_lists.is_first_time_user AS is_first_time_order_user,
transaction_orders_lists.is_first_time_user_completed AS is_first_time_order_user_completed,
transaction_orders_lists.base_to_quote_currency_exchange_rate AS order_base_to_quote_currency_exchange_rate ,
transaction_orders_lists.base_value AS order_base_value ,
transaction_orders_lists.base_service_charge AS order_base_service_charge ,
transaction_orders_lists.base_tax AS order_base_tax ,
transaction_orders_lists.quote_value AS order_quote_value ,
transaction_orders_lists.value_rm AS order_value_rm ,
transaction_orders_lists.service_charge_rm AS order_service_charge_rm ,
transaction_orders_lists.tax_rm AS order_tax_rm ,
transaction_orders_lists.total_value_rm AS total_order_value_rm,
transaction_orders_lists.order_created_datetime,
transaction_orders_lists.order_pending_submission_datetime,
transaction_orders_lists.order_pending_verification_datetime,
transaction_orders_lists.order_approved_datetime,
transaction_orders_lists.order_completed_datetime,
transaction_orders_lists.order_rejected_datetime,
transaction_orders_lists.order_suspended_datetime,
transaction_orders_lists.order_expired_datetime,
transaction_orders_lists.next_order_created_datetime,
transaction_orders_lists.order_expected_expired_datetime,
transaction_costs_lists.transaction_cost_id AS cost_id,
transaction_costs_lists.bank_id AS currency_vendor_bank_id,
transaction_costs_lists.supplier_company_id AS supplier_company_id,
transaction_costs_lists.transaction_type AS cost_transaction_type,
transaction_costs_lists.payment_method AS cost_payment_method,
transaction_costs_lists.status AS cost_status,
transaction_costs_lists.payment_reference AS cost_payment_reference,
transaction_costs_lists.bill_number AS cost_bill_number,
transaction_costs_lists.base_value AS cost_base_value,
transaction_costs_lists.quote_value AS cost_quote_value,
transaction_costs_lists.base_to_quote_currency_exchange_rate AS cost_base_to_quote_currency_exchange_rate,
transaction_costs_lists.base_tax AS cost_base_tax,
transaction_costs_lists.base_service_charge AS cost_base_service_charge,
transaction_costs_lists.value_rm AS cost_value_rm,
transaction_costs_lists.tax_rm AS cost_tax_rm,
transaction_costs_lists.service_charge_rm AS cost_service_charge_rm,
transaction_costs_lists.total_value_rm AS total_cost_value_rm,
transaction_costs_lists.cost_created_datetime,
transaction_costs_lists.cost_pending_submission_datetime,
transaction_costs_lists.cost_pending_verification_datetime,
transaction_costs_lists.cost_approved_datetime,
transaction_costs_lists.cost_completed_datetime,
transaction_costs_lists.cost_rejected_datetime,
'{{ modules.datetime.datetime.now(modules.pytz.timezone("Asia/Kuala_Lumpur")) }}' AS _dbt_ran_datetime
FROM transaction_orders_lists
LEFT JOIN transaction_costs_lists
ON transaction_orders_lists.transaction_order_id = transaction_costs_lists.transaction_order_id
LEFT JOIN bookings_lists
ON transaction_orders_lists.booking_id = bookings_lists.booking_id
),
-- FINAL
final__fct_exchange__transaction_orders AS (
final_fct_exchange__new_orders AS (
SELECT
-- ids
transaction_order_id,
order_id,
booking_id,
booking_marking_id,
company_id,
user_id,
bank_id,
base_currency_id,
cost_id,
supplier_company_id,
currency_vendor_bank_id,
fix_currency_id,
quote_currency_id,
base_currency_id,
-- dimensions
transaction_type,
payment_method,
service_type,
booking_status,
is_first_time_booking_company,
is_first_time_booking_company_completed,
is_first_time_booking_user,
is_first_time_booking_user_completed,
is_first_time_order_company,
is_first_time_order_company_completed,
is_first_time_order_user,
is_first_time_order_user_completed,
order_transaction_type,
order_payment_method,
order_status,
payment_reference,
bill_number,
is_first_time_company,
is_first_time_company_completed,
is_first_time_user,
is_first_time_user_completed,
order_payment_reference,
order_bill_number,
cost_transaction_type,
cost_payment_method,
cost_status,
cost_payment_reference,
cost_bill_number,
-- measures
base_to_quote_currency_exchange_rate,
base_value,
base_service_charge,
base_tax,
quote_value,
value_rm,
service_charge_rm,
tax_rm,
total_value_rm,
estimate_booking_base_to_quote_currency_exchange_rate,
estimate_booking_quote_value,
estimate_booking_base_value,
estimate_booking_value_rm,
order_base_to_quote_currency_exchange_rate,
order_base_value,
order_base_service_charge,
order_base_tax,
order_quote_value,
order_value_rm,
order_service_charge_rm,
order_tax_rm,
total_order_value_rm,
cost_base_value,
cost_quote_value,
cost_base_to_quote_currency_exchange_rate,
cost_base_tax,
cost_base_service_charge,
cost_value_rm,
cost_tax_rm,
cost_service_charge_rm,
total_cost_value_rm,
-- date/times
order_expired_datetime,
order_deleted_datetime,
order_updated_datetime,
booking_created_datetime,
booking_approved_datetime,
booking_completed_datetime,
booking_suspended_datetime,
order_created_datetime,
order_expected_expired_datetime,
order_pending_submission_datetime,
order_pending_verification_datetime,
order_approved_datetime,
order_completed_datetime,
order_rejected_datetime,
order_suspended_datetime,
order_expired_datetime,
next_order_created_datetime,
cost_created_datetime,
cost_pending_submission_datetime,
cost_pending_verification_datetime,
cost_approved_datetime,
cost_completed_datetime,
cost_rejected_datetime,
-- metadata
_dbt_ran_datetime
FROM transaction_orders_lists
FROM bookings_orders_costs_join
)
SELECT * FROM final__fct_exchange__transaction_orders
SELECT * FROM final_fct_exchange__new_orders