Code does Not Read in FY 2025 Data
1. The problem is the code below reads only data up until 2024, although
the table has been updated with latest data that contains 2025, for some
odd reason it is not pulling in or showing data when filtering for 2025 or
even 2024 and later, which should contain all the latest data.
2. SQL: "SET DATEFIRST 7;-- Holding Period Query
SELECT
3.
-- Holding Period Query
SELECT
WEEK_NUMBER."Week Number",
WM_GH."Calendar day",
WM_GH."Month/Week",
WM_GH."Day name",
WM_GH."Company",
WM_GH."Material Code",
WM_GH."Metric",
WM_GH."Metric Value"
FROM
(
SELECT
"Calendar day",
"Month/Week",
"Day name",
"Company",
"Material Code",
"Metric",
"Metric Value"
FROM
"Prod"."IMETA_Weekly_Metrics_in_Focus_Global_Stock_View_SAP_BW" AS WM_GH
WHERE
WM_GH."Metric Focus" LIKE 'Weekly'
AND WM_GH."Calendar day" <= (
SELECT MAX(WEEK_NUMBER."Date") AS "MAX DATE"
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date") AS
"Week number"
FROM
"dbo"."IMETA_Calendar_Days_Data_Table_Copy10" AS WEEK_NUMBER
WHERE
"FY" IS NOT NULL AND "FY" >= 'FY24'
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
AND "Number of days" = "Day number in weeks"
)
) AS WM_GH
LEFT OUTER JOIN
(
SELECT
*
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week number" ORDER BY
"Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week number" ORDER BY
"Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date") AS "Week
number"
FROM
"dbo"."IMETA_Calendar_Days_Data_Table_Copy10" AS
WEEK_NUMBER
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
) AS WEEK_NUMBER
ON
WM_GH."Calendar day" = WEEK_NUMBER."Date"
ORDER BY
WM_GH."Calendar day" DESC;
Attachments:
On 7/16/24 05:53, Anthony Apollis wrote:
1. The problem is the code below reads only data up until 2024,
although the table has been updated with latest data that contains
2025, for some odd reason it is not pulling in or showing data when
filtering for 2025 or even 2024 and later, which should contain all
the latest data.
2. SQL: "SET DATEFIRST 7;-- Holding Period Query
I don't recognize SET DATEFIRST 7 as Postgres command.
What database are you running this on?
Have not gone through the below fully, but this:
"FY" IS NOT NULL AND "FY" >= 'FY24'
and this
WEEK_NUMBER."Date" < CURRENT_DATE
would seem to be at odds if you want to reach into 2025.
SELECT
3.
-- Holding Period Query
SELECT
WEEK_NUMBER."Week Number",
WM_GH."Calendar day",
WM_GH."Month/Week",
WM_GH."Day name",
WM_GH."Company",
WM_GH."Material Code",
WM_GH."Metric",
WM_GH."Metric Value"
FROM
(
SELECT
"Calendar day",
"Month/Week",
"Day name",
"Company",
"Material Code",
"Metric",
"Metric Value"
FROM"Prod"."IMETA_Weekly_Metrics_in_Focus_Global_Stock_View_SAP_BW" AS WM_GH
WHERE
WM_GH."Metric Focus" LIKE 'Weekly'
AND WM_GH."Calendar day" <= (
SELECT MAX(WEEK_NUMBER."Date") AS "MAX DATE"
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date")
AS "Week number"
FROM"dbo"."IMETA_Calendar_Days_Data_Table_Copy10" AS WEEK_NUMBER
WHERE
"FY" IS NOT NULL AND "FY" >= 'FY24'
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
AND "Number of days" = "Day number in weeks"
)
) AS WM_GH
LEFT OUTER JOIN
(
SELECT
*
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date") AS "Week
number"
FROM
"dbo"."IMETA_Calendar_Days_Data_Table_Copy10"
AS WEEK_NUMBER
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
) AS WEEK_NUMBER
ON
WM_GH."Calendar day" = WEEK_NUMBER."Date"
ORDER BY
WM_GH."Calendar day" DESC;
--
Adrian Klaver
adrian.klaver@aklaver.com
I am using Postgres and SQL Server.
Can you test the data pls.
On Tue, 16 Jul 2024 at 16:45, Adrian Klaver <adrian.klaver@aklaver.com>
wrote:
Show quoted text
On 7/16/24 05:53, Anthony Apollis wrote:
1. The problem is the code below reads only data up until 2024,
although the table has been updated with latest data that contains
2025, for some odd reason it is not pulling in or showing data when
filtering for 2025 or even 2024 and later, which should contain all
the latest data.
2. SQL: "SET DATEFIRST 7;-- Holding Period QueryI don't recognize SET DATEFIRST 7 as Postgres command.
What database are you running this on?
Have not gone through the below fully, but this:
"FY" IS NOT NULL AND "FY" >= 'FY24'
and this
WEEK_NUMBER."Date" < CURRENT_DATE
would seem to be at odds if you want to reach into 2025.
SELECT
3.
-- Holding Period Query
SELECT
WEEK_NUMBER."Week Number",
WM_GH."Calendar day",
WM_GH."Month/Week",
WM_GH."Day name",
WM_GH."Company",
WM_GH."Material Code",
WM_GH."Metric",
WM_GH."Metric Value"
FROM
(
SELECT
"Calendar day",
"Month/Week",
"Day name",
"Company",
"Material Code",
"Metric",
"Metric Value"
FROM"Prod"."IMETA_Weekly_Metrics_in_Focus_Global_Stock_View_SAP_BW" AS
WM_GH
WHERE
WM_GH."Metric Focus" LIKE 'Weekly'
AND WM_GH."Calendar day" <= (
SELECT MAX(WEEK_NUMBER."Date") AS "MAX DATE"
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date")
AS "Week number"
FROM"dbo"."IMETA_Calendar_Days_Data_Table_Copy10" AS WEEK_NUMBER
WHERE
"FY" IS NOT NULL AND "FY" >= 'FY24'
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
AND "Number of days" = "Day number in weeks"
)
) AS WM_GH
LEFT OUTER JOIN
(
SELECT
*
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date") AS "Week
number"
FROM
"dbo"."IMETA_Calendar_Days_Data_Table_Copy10"
AS WEEK_NUMBER
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
) AS WEEK_NUMBER
ON
WM_GH."Calendar day" = WEEK_NUMBER."Date"
ORDER BY
WM_GH."Calendar day" DESC;--
Adrian Klaver
adrian.klaver@aklaver.com
On 16/07/2024 15:59, Anthony Apollis wrote:
I am using Postgres and SQL Server.
Can you test the data pls.
Adrian has already pointed out a possible cause of the issue you are
seeing - you should test further based on this.
Ray.
--
Raymond O'Donnell // Galway // Ireland
ray@rodonnell.ie
On 7/16/24 07:59, Anthony Apollis wrote:
I am using Postgres and SQL Server.
Does the query 'fail' on both platforms?
Can you test the data pls.
Don't know what I am supposed to see?
You will need to provide an explanation of what you want the query to
return vs what is currently being returned. Probably easiest to provide
a sample of data and what you expect the query to return from that data.
Do this inline in your post.
FYI, further investigation of the zipped data shows the 2025 data has FY
values of NULL which would fail "FY" IS NOT NULL.
On Tue, 16 Jul 2024 at 16:45, Adrian Klaver <adrian.klaver@aklaver.com
<mailto:adrian.klaver@aklaver.com>> wrote:On 7/16/24 05:53, Anthony Apollis wrote:
1. The problem is the code below reads only data up until 2024,
although the table has been updated with latest data thatcontains
2025, for some odd reason it is not pulling in or showing
data when
filtering for 2025 or even 2024 and later, which should
contain all
the latest data.
2. SQL: "SET DATEFIRST 7;-- Holding Period QueryI don't recognize SET DATEFIRST 7 as Postgres command.
What database are you running this on?
Have not gone through the below fully, but this:
"FY" IS NOT NULL AND "FY" >= 'FY24'
and this
WEEK_NUMBER."Date" < CURRENT_DATE
would seem to be at odds if you want to reach into 2025.
SELECT
3.
-- Holding Period Query
SELECT
WEEK_NUMBER."Week Number",
WM_GH."Calendar day",
WM_GH."Month/Week",
WM_GH."Day name",
WM_GH."Company",
WM_GH."Material Code",
WM_GH."Metric",
WM_GH."Metric Value"
FROM
(
SELECT
"Calendar day",
"Month/Week",
"Day name",
"Company",
"Material Code",
"Metric",
"Metric Value"
FROM"Prod"."IMETA_Weekly_Metrics_in_Focus_Global_Stock_View_SAP_BW" AS
WM_GHWHERE
WM_GH."Metric Focus" LIKE 'Weekly'
AND WM_GH."Calendar day" <= (
SELECT MAX(WEEK_NUMBER."Date") AS "MAX DATE"
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week
number" ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROMWEEK_NUMBER."Date")
AS "Week number"
FROM"dbo"."IMETA_Calendar_Days_Data_Table_Copy10" AS WEEK_NUMBER
WHERE
"FY" IS NOT NULL AND "FY" >= 'FY24'
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
AND "Number of days" = "Day number in weeks"
)
) AS WM_GH
LEFT OUTER JOIN
(
SELECT
*
FROM (
SELECT
"Date",
"Week number",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Week number") AS "Number of days",
COUNT("Date") OVER (PARTITION BY "Week number"
ORDER BY "Date") AS "Day number in weeks"
FROM (
SELECT DISTINCT
"Date",
EXTRACT(WEEK FROM WEEK_NUMBER."Date") AS"Week
number"
FROM"dbo"."IMETA_Calendar_Days_Data_Table_Copy10"
AS WEEK_NUMBER
) AS W_MAX
) AS WEEK_NUMBER
WHERE
WEEK_NUMBER."Date" < CURRENT_DATE
) AS WEEK_NUMBER
ON
WM_GH."Calendar day" = WEEK_NUMBER."Date"
ORDER BY
WM_GH."Calendar day" DESC;--
Adrian Klaver
adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>
--
Adrian Klaver
adrian.klaver@aklaver.com
On 7/16/24 08:15, Anthony Apollis wrote:
Reply to list also
Ccing list
i did attached the sample data.
Yes I know and the data for 2025 will fail because they are like this:
Date FY Period Quarter Day Month Year Loaddate
2025-01-10 00:00:00.000 NULL NULL NULL 10 1 2025 2023-07-10 11:55:09.733
The FY values are NULL and they will not be found by:
"FY" IS NOT NULL AND "FY" >= 'FY24'
For the updated data that is for 2026 and 2027 there are FY values, but
then you do:
WEEK_NUMBER."Date" < CURRENT_DATE
where WEEK_NUMBER is the alias for the sub-select that filtered on:
"FY" IS NOT NULL AND "FY" >= 'FY24'
So even if the sub-select found values in the future you filter them out
with < CURRENT_DATE
--
Adrian Klaver
adrian.klaver@aklaver.com
Import Notes
Reply to msg id not found: CAJyMCYJSnh8HvWumnd18uzLtU+ZjimF+DeSaBV=uXVq8HbFEzA@mail.gmail.com
On 7/16/24 08:26, Anthony Apollis wrote:
Again reply to list also
Ccing list
I included a bigger data set. This code only pics up data up until 2024
in new table which has data for 2025. Something with the Maxdate
calculation is wrong.
-- Step 1: Define the Fiscal Calendar
WITH FiscalCalendar AS (
SELECT 'FY25' AS FY, 'P1' AS Period, '2024-07-01'::date AS
PeriodStart, '2024-08-03'::date AS PeriodEnd
UNION ALL
SELECT 'FY25', 'P2', '2024-08-04'::date, '2024-08-31'::date
UNION ALL
SELECT 'FY25', 'P3', '2024-09-01'::date, '2024-09-28'::date
UNION ALL
SELECT 'FY25', 'P4', '2024-09-29'::date, '2024-11-02'::date
UNION ALL
SELECT 'FY25', 'P5', '2024-11-03'::date, '2024-11-30'::date
UNION ALL
SELECT 'FY25', 'P6', '2024-12-01'::date, '2024-12-31'::date
UNION ALL
SELECT 'FY25', 'P7', '2025-01-01'::date, '2025-02-01'::date
UNION ALL
SELECT 'FY25', 'P8', '2025-02-02'::date, '2025-03-01'::date
UNION ALL
SELECT 'FY25', 'P9', '2025-03-02'::date, '2025-03-29'::date
UNION ALL
SELECT 'FY25', 'P10', '2025-03-30'::date, '2025-05-03'::date
UNION ALL
SELECT 'FY25', 'P11', '2025-05-04'::date, '2025-05-31'::date
UNION ALL
SELECT 'FY25', 'P12', '2025-06-01'::date, '2025-06-30'::date
),-- Step 2: Get Calendar Data with full week details
CalendarData AS (
SELECT DISTINCT
c."Date",
EXTRACT(WEEK FROM c."Date") AS "WeekNumber",
f.PeriodEnd,
COUNT(c."Date") OVER (PARTITION BY EXTRACT(WEEK FROM c."Date")
ORDER BY c."Date") AS "NumberOfDays",
ROW_NUMBER() OVER (PARTITION BY EXTRACT(WEEK FROM c."Date")
ORDER BY c."Date") AS "DayNumberInWeek"
FROM
"Prod"."IMETA - Calendar Days Data_Table_Temp_Copy" c
INNER JOIN
FiscalCalendar f ON c."Date" BETWEEN f.PeriodStart AND
f.PeriodEnd AND c."FY" = f.FY
WHERE
c."FY" = 'FY25'
),-- Step 3: Filter to get full weeks only
FullWeeks AS (
SELECT
"Date",
"WeekNumber",
"PeriodEnd",
"NumberOfDays",
"DayNumberInWeek"
FROM
CalendarData
WHERE
"NumberOfDays" = 7
),-- Step 4: Get the maximum date from full weeks that is before today
MaxDate AS (
SELECT
MAX("Date") AS "MaxDate"
FROM
FullWeeks
WHERE
"Date" < CURRENT_DATE
)-- Step 5: Select the maximum date
SELECT
"MaxDate"
FROM
MaxDate;On Tue, 16 Jul 2024 at 17:13, Adrian Klaver <adrian.klaver@aklaver.com
<mailto:adrian.klaver@aklaver.com>> wrote:On 7/16/24 07:59, Anthony Apollis wrote:
I am using Postgres and SQL Server.
Does the query 'fail' on both platforms?
Can you test the data pls.
Don't know what I am supposed to see?
You will need to provide an explanation of what you want the query to
return vs what is currently being returned. Probably easiest to provide
a sample of data and what you expect the query to return from that
data.
Do this inline in your post.FYI, further investigation of the zipped data shows the 2025 data
has FY
values of NULL which would fail "FY" IS NOT NULL.On Tue, 16 Jul 2024 at 16:45, Adrian Klaver
<adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>
<mailto:adrian.klaver@aklaver.com
<mailto:adrian.klaver@aklaver.com>>> wrote:
On 7/16/24 05:53, Anthony Apollis wrote:
> 1. The problem is the code below reads only data up until2024,
> although the table has been updated with latest data that
contains
> 2025, for some odd reason it is not pulling in or showing
data when
> filtering for 2025 or even 2024 and later, which should
contain all
> the latest data.
> 2. SQL: "SET DATEFIRST 7;-- Holding Period QueryI don't recognize SET DATEFIRST 7 as Postgres command.
What database are you running this on?
Have not gone through the below fully, but this:
"FY" IS NOT NULL AND "FY" >= 'FY24'
and this
WEEK_NUMBER."Date" < CURRENT_DATE
would seem to be at odds if you want to reach into 2025.
>
> SELECT
>
> 3.
>
> -- Holding Period Query
> SELECT
> WEEK_NUMBER."Week Number",
> WM_GH."Calendar day",
> WM_GH."Month/Week",
> WM_GH."Day name",
> WM_GH."Company",
> WM_GH."Material Code",
> WM_GH."Metric",
> WM_GH."Metric Value"
> FROM
> (
> SELECT
> "Calendar day",
> "Month/Week",
> "Day name",
> "Company",
> "Material Code",
> "Metric",
> "Metric Value"
> FROM
>
>"Prod"."IMETA_Weekly_Metrics_in_Focus_Global_Stock_View_SAP_BW" AS
WM_GH
> WHERE
> WM_GH."Metric Focus" LIKE 'Weekly'
> AND WM_GH."Calendar day" <= (
> SELECT MAX(WEEK_NUMBER."Date") AS"MAX DATE"
> FROM (
> SELECT
> "Date",
> "Week number",
> COUNT("Date") OVER (PARTITIONBY "Week
> number" ORDER BY "Week number") AS "Number of days",
> COUNT("Date") OVER (PARTITIONBY "Week
> number" ORDER BY "Date") AS "Day number in weeks"
> FROM (
> SELECT DISTINCT
> "Date",
> EXTRACT(WEEK FROM
WEEK_NUMBER."Date")
> AS "Week number"
> FROM
>
> "dbo"."IMETA_Calendar_Days_Data_Table_Copy10" ASWEEK_NUMBER
> WHERE
> "FY" IS NOT NULL AND "FY"
= 'FY24'
> ) AS W_MAX
> ) AS WEEK_NUMBER
> WHERE
> WEEK_NUMBER."Date" < CURRENT_DATE
> AND "Number of days" = "Daynumber in weeks"
> )
> ) AS WM_GH
> LEFT OUTER JOIN
> (
> SELECT
> *
> FROM (
> SELECT
> "Date",
> "Week number",
> COUNT("Date") OVER (PARTITION BY"Week number"
> ORDER BY "Week number") AS "Number of days",
> COUNT("Date") OVER (PARTITION BY"Week number"
> ORDER BY "Date") AS "Day number in weeks"
> FROM (
> SELECT DISTINCT
> "Date",
> EXTRACT(WEEK FROMWEEK_NUMBER."Date") AS
"Week
> number"
> FROM
>
"dbo"."IMETA_Calendar_Days_Data_Table_Copy10"
> AS WEEK_NUMBER
> ) AS W_MAX
> ) AS WEEK_NUMBER
> WHERE
> WEEK_NUMBER."Date" < CURRENT_DATE
> ) AS WEEK_NUMBER
> ON
> WM_GH."Calendar day" = WEEK_NUMBER."Date"
> ORDER BY
> WM_GH."Calendar day" DESC;
>--
Adrian Klaver
adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com><mailto:adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>>
--
Adrian Klaver
adrian.klaver@aklaver.com <mailto:adrian.klaver@aklaver.com>
--
Adrian Klaver
adrian.klaver@aklaver.com
Import Notes
Reply to msg id not found: CAJyMCY+CqLtkETj88qAEX=HGTyfsND3CGpvzEMcLRFVcUNEJSQ@mail.gmail.com
Only data up until 2024 is picked up in Revised table, whic contains 2025
data. THe Maxdate calculation seems to be the problem.
-- Step 1: Define the Fiscal Calendar
WITH FiscalCalendar AS (
SELECT 'FY25' AS FY, 'P1' AS Period, '2024-07-01'::date AS PeriodStart,
'2024-08-03'::date AS PeriodEnd
UNION ALL
SELECT 'FY25', 'P2', '2024-08-04'::date, '2024-08-31'::date
UNION ALL
SELECT 'FY25', 'P3', '2024-09-01'::date, '2024-09-28'::date
UNION ALL
SELECT 'FY25', 'P4', '2024-09-29'::date, '2024-11-02'::date
UNION ALL
SELECT 'FY25', 'P5', '2024-11-03'::date, '2024-11-30'::date
UNION ALL
SELECT 'FY25', 'P6', '2024-12-01'::date, '2024-12-31'::date
UNION ALL
SELECT 'FY25', 'P7', '2025-01-01'::date, '2025-02-01'::date
UNION ALL
SELECT 'FY25', 'P8', '2025-02-02'::date, '2025-03-01'::date
UNION ALL
SELECT 'FY25', 'P9', '2025-03-02'::date, '2025-03-29'::date
UNION ALL
SELECT 'FY25', 'P10', '2025-03-30'::date, '2025-05-03'::date
UNION ALL
SELECT 'FY25', 'P11', '2025-05-04'::date, '2025-05-31'::date
UNION ALL
SELECT 'FY25', 'P12', '2025-06-01'::date, '2025-06-30'::date
),
-- Step 2: Get Calendar Data with full week details
CalendarData AS (
SELECT DISTINCT
c."Date",
EXTRACT(WEEK FROM c."Date") AS "WeekNumber",
f.PeriodEnd,
COUNT(c."Date") OVER (PARTITION BY EXTRACT(WEEK FROM c."Date")
ORDER BY c."Date") AS "NumberOfDays",
ROW_NUMBER() OVER (PARTITION BY EXTRACT(WEEK FROM c."Date") ORDER
BY c."Date") AS "DayNumberInWeek"
FROM
"Prod"."IMETA - Calendar Days Data_Table_Temp_Copy" c
INNER JOIN
FiscalCalendar f ON c."Date" BETWEEN f.PeriodStart AND f.PeriodEnd
AND c."FY" = f.FY
WHERE
c."FY" = 'FY25'
),
-- Step 3: Filter to get full weeks only
FullWeeks AS (
SELECT
"Date",
"WeekNumber",
"PeriodEnd",
"NumberOfDays",
"DayNumberInWeek"
FROM
CalendarData
WHERE
"NumberOfDays" = 7
),
-- Step 4: Get the maximum date from full weeks that is before today
MaxDate AS (
SELECT
MAX("Date") AS "MaxDate"
FROM
FullWeeks
WHERE
"Date" < CURRENT_DATE
)
-- Step 5: Select the maximum date
SELECT
"MaxDate"
FROM
MaxDate;
On Tue, 16 Jul 2024 at 17:26, Adrian Klaver <adrian.klaver@aklaver.com>
wrote:
Show quoted text
On 7/16/24 08:15, Anthony Apollis wrote:
Reply to list also
Ccing listi did attached the sample data.
Yes I know and the data for 2025 will fail because they are like this:
Date FY Period Quarter Day Month Year Loaddate
2025-01-10 00:00:00.000 NULL NULL NULL 10 1 2025
2023-07-10 11:55:09.733The FY values are NULL and they will not be found by:
"FY" IS NOT NULL AND "FY" >= 'FY24'
For the updated data that is for 2026 and 2027 there are FY values, but
then you do:WEEK_NUMBER."Date" < CURRENT_DATE
where WEEK_NUMBER is the alias for the sub-select that filtered on:
"FY" IS NOT NULL AND "FY" >= 'FY24'
So even if the sub-select found values in the future you filter them out
with < CURRENT_DATE--
Adrian Klaver
adrian.klaver@aklaver.com
The Calendar Tables should adhere to this business rule/calendar. Original
Table seem to be correct.
Reporting Month FY22 FY23 FY24 FY25 FY26 FY27
Period End Calendar Days Period End Calendar Days Period End Calendar
Days Period
End Calendar Days Period End Calendar Days Period End Calendar Days
P1 31-Jul 2021 (Sat) 31 30-Jul 2022 (Sat) 30 05-Aug 2023 (Sat) 36 03-Aug
2024 (Sat) 34 02-Aug 2025 (Sat) 33 01-Aug 2026 (Sat) 32
P2 28-Aug 2021 (Sat) 28 27-Aug 2022 (Sat) 28 02-Sep 2023 (Sat) 28 31-Aug
2024 (Sat) 28 30-Aug 2025 (Sat) 28 29-Aug 2026 (Sat) 28
P3 25-Sep 2021 (Sat) 28 24-Sep 2022 (Sat) 28 30-Sep 2023 (Sat) 28 28-Sep
2024 (Sat) 28 27-Sep 2025 (Sat) 28 26-Sep 2026 (Sat) 28
P4 30-Oct 2021 (Sat) 35 29-Oct 2022 (Sat) 35 04-Nov 2023 (Sat) 35 02-Nov
2024 (Sat) 35 01-Nov 2025 (Sat) 35 31-Oct 2026 (Sat) 35
P5 27-Nov 2021 (Sat) 28 26-Nov 2022 (Sat) 28 02-Dec 2023 (Sat) 28 30-Nov
2024 (Sat) 28 29-Nov 2025 (Sat) 28 28-Nov 2026 (Sat) 28
P6 31-Dec 2021 (Fri) 34 31-Dec 2022 (Sat) 35 31-Dec 2023 (Sun) 29 31-Dec
2024 (Tue) 31 31-Dec 2025 (Wed) 32 31-Dec 2026 (Thu) 33
P7 29-Jan 2022 (Sat) 29 04-Feb 2023 (Sat) 35 03-Feb 2024 (Sat) 34 01-Feb
2025 (Sat) 32 31-Jan 2026 (Sat) 31 30-Jan 2027 (Sat) 30
P8 26-Feb 2022 (Sat) 28 04-Mar 2023 (Sat) 28 02-Mar 2024 (Sat) 28 01-Mar
2025 (Sat) 28 28-Feb 2026 (Sat) 28 27-Feb 2027 (Sat) 28
P9 26-Mar 2022 (Sat) 28 01-Apr 2023 (Sat) 28 30-Mar 2024 (Sat) 28 29-Mar
2025 (Sat) 28 28-Mar 2026 (Sat) 28 27-Mar 2027 (Sat) 28
P10 30-Apr 2022 (Sat) 35 06-May 2023 (Sat) 35 04-May 2024 (Sat) 35 03-May
2025 (Sat) 35 02-May 2026 (Sat) 35 01-May 2027 (Sat) 35
P11 28-May 2022 (Sat) 28 03-Jun 2023 (Sat) 28 01-Jun 2024 (Sat) 28 31-May
2025 (Sat) 28 30-May 2026 (Sat) 28 29-May 2027 (Sat) 28
P12 30-Jun 2022 (Thu) 33 30-Jun 2023 (Fri) 27 30-Jun 2024 (Sun) 29 30-Jun
2025 (Mon) 30 30-Jun 2026 (Tue) 31 30-Jun 2027 (Wed) 32
Total Fiscal Year Days 365 365 366 365 365 365
Day 1 of Fiscal year: 30-Jun 2021 (Wed) 30-Jun 2022 (Thu) 30-Jun 2023
(Fri) 30-Jun 2024 (Sun) 30-Jun 2025 (Mon) 30-Jun 2026 (Tue)
On Tue, 16 Jul 2024 at 17:28, Anthony Apollis <anthony.apollis@gmail.com>
wrote:
Show quoted text
Only data up until 2024 is picked up in Revised table, whic contains 2025
data. THe Maxdate calculation seems to be the problem.
-- Step 1: Define the Fiscal Calendar
WITH FiscalCalendar AS (
SELECT 'FY25' AS FY, 'P1' AS Period, '2024-07-01'::date AS
PeriodStart, '2024-08-03'::date AS PeriodEnd
UNION ALL
SELECT 'FY25', 'P2', '2024-08-04'::date, '2024-08-31'::date
UNION ALL
SELECT 'FY25', 'P3', '2024-09-01'::date, '2024-09-28'::date
UNION ALL
SELECT 'FY25', 'P4', '2024-09-29'::date, '2024-11-02'::date
UNION ALL
SELECT 'FY25', 'P5', '2024-11-03'::date, '2024-11-30'::date
UNION ALL
SELECT 'FY25', 'P6', '2024-12-01'::date, '2024-12-31'::date
UNION ALL
SELECT 'FY25', 'P7', '2025-01-01'::date, '2025-02-01'::date
UNION ALL
SELECT 'FY25', 'P8', '2025-02-02'::date, '2025-03-01'::date
UNION ALL
SELECT 'FY25', 'P9', '2025-03-02'::date, '2025-03-29'::date
UNION ALL
SELECT 'FY25', 'P10', '2025-03-30'::date, '2025-05-03'::date
UNION ALL
SELECT 'FY25', 'P11', '2025-05-04'::date, '2025-05-31'::date
UNION ALL
SELECT 'FY25', 'P12', '2025-06-01'::date, '2025-06-30'::date
),-- Step 2: Get Calendar Data with full week details
CalendarData AS (
SELECT DISTINCT
c."Date",
EXTRACT(WEEK FROM c."Date") AS "WeekNumber",
f.PeriodEnd,
COUNT(c."Date") OVER (PARTITION BY EXTRACT(WEEK FROM c."Date")
ORDER BY c."Date") AS "NumberOfDays",
ROW_NUMBER() OVER (PARTITION BY EXTRACT(WEEK FROM c."Date") ORDER
BY c."Date") AS "DayNumberInWeek"
FROM
"Prod"."IMETA - Calendar Days Data_Table_Temp_Copy" c
INNER JOIN
FiscalCalendar f ON c."Date" BETWEEN f.PeriodStart AND f.PeriodEnd
AND c."FY" = f.FY
WHERE
c."FY" = 'FY25'
),-- Step 3: Filter to get full weeks only
FullWeeks AS (
SELECT
"Date",
"WeekNumber",
"PeriodEnd",
"NumberOfDays",
"DayNumberInWeek"
FROM
CalendarData
WHERE
"NumberOfDays" = 7
),-- Step 4: Get the maximum date from full weeks that is before today
MaxDate AS (
SELECT
MAX("Date") AS "MaxDate"
FROM
FullWeeks
WHERE
"Date" < CURRENT_DATE
)-- Step 5: Select the maximum date
SELECT
"MaxDate"
FROM
MaxDate;On Tue, 16 Jul 2024 at 17:26, Adrian Klaver <adrian.klaver@aklaver.com>
wrote:On 7/16/24 08:15, Anthony Apollis wrote:
Reply to list also
Ccing listi did attached the sample data.
Yes I know and the data for 2025 will fail because they are like this:
Date FY Period Quarter Day Month Year Loaddate
2025-01-10 00:00:00.000 NULL NULL NULL 10 1 2025
2023-07-10 11:55:09.733The FY values are NULL and they will not be found by:
"FY" IS NOT NULL AND "FY" >= 'FY24'
For the updated data that is for 2026 and 2027 there are FY values, but
then you do:WEEK_NUMBER."Date" < CURRENT_DATE
where WEEK_NUMBER is the alias for the sub-select that filtered on:
"FY" IS NOT NULL AND "FY" >= 'FY24'
So even if the sub-select found values in the future you filter them out
with < CURRENT_DATE--
Adrian Klaver
adrian.klaver@aklaver.com
On 7/16/24 08:28, Anthony Apollis wrote:
Only data up until 2024 is picked up in Revised table, whic contains
2025 data. THe Maxdate calculation seems to be the problem.
This:
'Only data up until 2024 is picked up in Revised table, whic contains
2025 data. .."
is consistent with:
MaxDate AS (
SELECT
MAX("Date") AS "MaxDate"
FROM
FullWeeks
WHERE
"Date" < CURRENT_DATE
)
I don't see how you expect to fetch data from the future when you limit
the data to the past.
--
Adrian Klaver
adrian.klaver@aklaver.com
That very same code is picking up all the data in the Original Table. Which
only contains past data.
On Tue, 16 Jul 2024 at 17:39, Adrian Klaver <adrian.klaver@aklaver.com>
wrote:
Show quoted text
On 7/16/24 08:28, Anthony Apollis wrote:
Only data up until 2024 is picked up in Revised table, whic contains
2025 data. THe Maxdate calculation seems to be the problem.This:
'Only data up until 2024 is picked up in Revised table, whic contains
2025 data. .."is consistent with:
MaxDate AS (
SELECT
MAX("Date") AS "MaxDate"
FROM
FullWeeks
WHERE
"Date" < CURRENT_DATE
)I don't see how you expect to fetch data from the future when you limit
the data to the past.--
Adrian Klaver
adrian.klaver@aklaver.com
On 7/16/24 08:57, Anthony Apollis wrote:
That very same code is picking up all the data in the Original Table.
Which only contains past data.
It would.
select CURRENT_DATE;
current_date
--------------
2024-07-16
"Date" < CURRENT_DATE
would return data < 2024-07-16.
You want data that has dates > 2024-07-16.
--
Adrian Klaver
adrian.klaver@aklaver.com
On Tue, Jul 16, 2024 at 7:59 AM Anthony Apollis <anthony.apollis@gmail.com>
wrote:
I am using Postgres and SQL Server.
Can you test the data pls.
Well, this is a PostgreSQL community so you should target it with your
communications.
If you want someone to actually test things here you probably will need to
produce a psql script that constructs the relevant database objects,
populates them, and then executes and produces the empty result you are
claiming you see.
In lieu of psql you could use an online fiddle to build this into a
self-contained executable form. Here is a link to one.
David J.