Showing posts with label PowerBI inventory analytics. Show all posts
Showing posts with label PowerBI inventory analytics. Show all posts

Wednesday, 6 March 2019

Inventory optimization and sales analytics using Microsoft PowerBI


End client problem:
 The end client is a Rs. 150(one hundred and fifty) crore rupee firm in the trading of steel tubes for casting purposes in the automotive sector.
Attached were the business issues for the end client:




·         The end client desires visibility into its inventory  in terms of the following parameters:
o   Ageing analysis
o   ABC categorization
o   Upper and lower control limit for inventory ageing




·         The end client data was in disparate sources of excel spreadsheets. The following were the issues with the excel spreadsheets:
o   The data in the spreadsheets was not clean
o   The end client business logic was to be coded.
o   There are multiple and disparate spreadsheets for stock, stock valuation and sales.
·         Code maintainability: the end client has limited skills on sql. The end client managers needed a system that could be maintained by its internal resources with ease and convenience. The end client requirements were as follows:
o   Maintaining of the business logic as per its requirements. This was specifically true for the inventory ageing analysis
o   Developing report and dashboards as per end client requirements.




In order to achieve the same, the etl team devised the following course of action.
·         Developed the solution using microsoft powerbi by using the same as an etl tool and a reporting tool.
·         The entire sql business logic of the end client was converted into microsoft powerbi etl logic.
·         The reports and dashboards were developed from the cleansed data of microsoft powerbi
·         Knowledge transfer was done for the end client to ensure smooth operations for the same.
Attached is the detailed description of the data cleansing, data auditing and data profiling for the end client.

The data providers given are as follows:









Attached are the data providers for the end client usage:
·         ‘a category ageing’: this gives a comprehensive analysis for ‘a’ category items in terms of abc analysis, ageing.
·         Abc category: this gives abc categorization by  weight
·         Ageing_new: this module was later developer to include ageing analysis by value
·         Final(2): sales analysis
·         Month_month_analysis:this is for sales analysis with built in month data.



Stock data:


This is the raw data given by end client.the raw data requires data cleansing.
the following were the errors marked for the end client:
1.)   Grade can be numeric as well as character.
2.)   Missing values of dim1, make, grade and dates. These details have to be marked to the end client. For example in dates, the end client could misrepresent the year of the date input.
Cleasing part:
Attached are the steps for cleansing:
·        

An addendum to grade of ‘-x’ was added. This was later removed in the powerbi.



Sales data:


·         Now the work for the etl team is to load the data in the powerbi.
Stock sheet:
Next step for the etl team :
·         Remove the extra variables from the data using powerbi.
·         It has to be done by the delimiter function.

A conditional column was added within powerbi for flat,rounds and sections business logic. Attached is the snapshot for the same:
The excess column of price was removed.

Next step:
·         Added custom column to find the value.
·         Value=wt*price
Attached is the snapshot for the same:
After that we have to do group by:
·         Grouped rows on the basis of grade,make,condition,section stock
·         Operation =sum




This is our required data for reporting.
Attached is snapshot for the same:






Next step  for ETL resource is to  do ABC categorization.
Steps for the abc categorization are as follows:
(the attached module sql code was converted into powerbi module for the same)
ALTER PROCEDURE [dbo].[SP_ABC_CATEGORIZATION_AGEING]
AS
BEGIN
TRUNCATE TABLE ABC_CATEGORIZATION_BBY;


WITH CTE1
AS(
SELECT A.*
--,B.CATEGORY

FROM BBY_STOCK_DEC A
--JOIN ABC_CATEGORIZATION1 B
-- ON A.GRADE=B.GRADE
-- AND A.CONDITION=B.CONDITION
-- AND A.MAKE=B.MAKE
-- AND A.DIM1=B.DIM1
WHERE D2 IS NULL

),
CTE2
AS(SELECT  DISTINCT  A.* FROM CTE1 A
JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.CONDITION=B.CONDITION
 AND A.MAKE=B.MAKE
 AND A.D1=B.D1),

 CTE2A
 AS(SELECT DISTINCT A.*,B.CATEGORY FROM CTE1 A
JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.CONDITION=B.CONDITION
 AND A.MAKE=B.MAKE
 AND A.D1=B.D1),

 CTE3
 AS(SELECT * FROM CTE1
 EXCEPT
 SELECT * FROM CTE2),
 /* PICK UP THAT STOCK WHERE CONDITION IS NULL AND MAKE IS NULL*/
 CTE4
 AS(SELECT DISTINCT A.* FROM CTE3 A
 JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE

 AND A.D1=B.D1
 WHERE A.CONDITION IS NULL
 AND A.MAKE IS NULL),
 CTE4A
 AS(SELECT DISTINCT A.*,B.CATEGORY FROM CTE3 A
 JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE

 AND A.D1=B.D1
 WHERE A.CONDITION IS NULL
 AND A.MAKE IS NULL),

 CTE5
 AS(SELECT * FROM CTE3
 EXCEPT
 SELECT * FROM CTE4),

 CTE6
 AS(SELECT DISTINCT A.* FROM CTE5 A
 JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.CONDITION=B.CONDITION

 AND A.D1=B.D1
 WHERE
A.MAKE IS NULL),
CTE6A
 AS(SELECT DISTINCT A.*,B.CATEGORY FROM CTE5 A
 JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.CONDITION=B.CONDITION

 AND A.D1=B.D1
 WHERE
A.MAKE IS NULL),
CTE7
AS(SELECT * FROM CTE5
EXCEPT
SELECT * FROM CTE6),
CTE8 AS(SELECT DISTINCT A.* FROM CTE7 A
JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.MAKE=B.MAKE

 AND A.D1=B.D1
 WHERE A.CONDITION IS NULL),
 CTE8A
 AS(SELECT DISTINCT  A.*,B.CATEGORY FROM CTE7 A
JOIN ABC_CATEGORIZATION B
 ON A.GRADE=B.GRADE
 AND A.MAKE=B.MAKE

 AND A.D1=B.D1
 WHERE A.CONDITION IS NULL),

 CTE9
 AS(SELECT * FROM CTE7
 EXCEPT
 SELECT * FROM CTE8),

 CTE9X
 AS(SELECT * FROM CTE9
 WHERE DATE IS NOT NULL),


 CTE9XA
 AS(SELECT DISTINCT  A.*,B.CATEGORY FROM CTE9X A
JOIN ABC_CATEGORIZATION B
 ON A.CONDITION=B.CONDITION
 AND A.MAKE=B.MAKE

 AND A.D1=B.D1
 WHERE A.GRADE IS NULL),

  CTE9XAB
 AS(SELECT DISTINCT  A.* FROM CTE9X A
JOIN ABC_CATEGORIZATION B
 ON A.CONDITION=B.CONDITION
 AND A.MAKE=B.MAKE

 AND A.D1=B.D1
 WHERE A.GRADE IS NULL),

 CTE9XC
 AS(SELECT * FROM CTE9X
 EXCEPT
 SELECT * FROM CTE9XAB),


 CTE9A
 AS(SELECT * FROM CTE2A
 UNION ALL

 SELECT * FROM CTE4A
 UNION ALL
 SELECT * FROM CTE6A
 UNION ALL

 SELECT * FROM CTE8A

 UNION ALL

 SELECT * FROM CTE9XA)



 SELECT * INTO #CTE9A FROM CTE9A;
 /*AGEING LOGIC AND TOTAL VALUE LOGIC*/
 WITH CTE10
 AS(SELECT A.*,
 DATEDIFF(DAY,ISNULL(CONVERT(DATETIME, DATE, 104),'2017-01-01 00:00:00.000'),GETDATE()) AS AGEING,
 CAST((A.BAL_WT)AS DECIMAL (18,2))* CAST((A.PRICE) AS DECIMAL(18,2)) AS TOTAL_VALUE
  FROM #CTE9A A)

 
  INSERT  INTO ABC_CATEGORIZATION_BBY
(
[DATE ]
      ,[LOT NO]
      ,[GRADE]
      ,[MAKE]
      ,[CONDITION]
      ,[D1]
      ,[D2]
      ,[BAL_WT]
      ,[PRICE]
      ,[CATEGORY]
      ,[AGEING]
      ,[TOTAL_VALUE]
)



      SELECT
     [DATE ]
      ,[LOT NO]
      ,[GRADE]
      ,[MAKE]
      ,[CONDITION]
      ,[D1]
      ,[D2]
      ,[BAL_WT]
      ,[PRICE]
      ,[CATEGORY]
      ,[AGEING]
      ,[TOTAL_VALUE]

FROM CTE10

END;

 


·         First step is to find cummulative sum.
o   CUMULATIVE_SUM_2 = CALCULATE(SUM(STOCK_VALUATION_TABLE[VALUE_SUM]),FILTER(ALL(STOCK_VALUATION_TABLE),STOCK_VALUATION_TABLE[Index]<=MAX(STOCK_VALUATION_TABLE[Index])))


·         Abc categorization:
ABC = IF(((STOCK_VALUATION_TABLE[CUMULATIVE_SUM_2]/SUM(STOCK_VALUATION_TABLE[VALUE_SUM]))*100)<=50, "A",IF(((STOCK_VALUATION_TABLE[CUMULATIVE_SUM_2]/SUM(STOCK_VALUATION_TABLE[VALUE_SUM]))*100)>50 && ((STOCK_VALUATION_TABLE[CUMULATIVE_SUM_2]/SUM(STOCK_VALUATION_TABLE[VALUE_SUM]))*100)<80,"B","C"))


Next step is the ageing work for the stock.
Ageing requires:
·         Date difference
o   date_diff = DATEDIFF(AGEING_NEW[Stock.Date],today(),DAY)

Attached is the ageing of the stock:
           

Formulae for ageing:
·         ageing = IF(AGEING_NEW[date_diff]>0 && AGEING_NEW[date_diff]<365,"LESS THAN 1 YEAR",IF(AGEING_NEW[date_diff]>365 && AGEING_NEW[date_diff]<730,"BETWEEN 1 AND 2 YEAR",IF(AGEING_NEW[date_diff]>730 && AGEING_NEW[date_diff] < 1095,"BETWEEN 2 AND 3 YEAR","MORE THAN 3 YEAR")))






 


 
  The final output for the end client was as follows:


·         ABC categorization in terms of stock value(sample attached below)

  •          Inventory ageing analysis in terms of weight(sample attached below)

  •       Month and month analysis(sample attached below):


Dashboard


Inventory  valuations from turbodata using DAX functionality
The etl team had the following target points for converting the sql code to dax for weighted average calculations.
·         Reduce the deployment times-remove the sql server extraction step
·         Reduce the execution times: remove the sql code execution steps
·         Handle large data loads: remove the bottlenecks associated with ms excel
·         High end data auditing  and data profiling solution using powerbi
·         Reduced costs of maintenance of the solution: reduced dependence on sql developers for the solution execution.

What this solution does not handle?
·         Incremental data loads from large disparate databases.
Attached document explains the conversion of the complex sql code to powerbi code
·         Summarization of the debit and credit entries: the etl team has assumed that all the debit entries and credit entries entailing the stock item details shall be considered irrespective of the voucher type details.
The attached sql code was converted into an aggregate using dax functionality
SELECT DISTINCT STOCKITEM_NM AS STOCK_ITEM_NAME INTO #DAYBOOK_STOCK_ITEM_NM FROM STG_ITEM;

CREATE INDEX #DAYBOOK_STOCK_ITEM_NM_IDX ON #DAYBOOK_STOCK_ITEM_NM
(STOCK_ITEM_NAME);
--DELETE FROM MOVING_AVERAGE_RATE WHERE ITEM_NAME IN (SELECT STOCK_ITEM_NAME FROM #DAYBOOK_STOCK_ITEM_NM);
WITH CTE1 AS (
                  SELECT E.DIM_COMPANY_ID,A.RATE,A.SALE_QUANTITY,(A.RATE*A.SALE_QUANTITY) AS SALE_VALUE,A.ORDER_DATE AS DATE_DESC,A.ITEM_SUPPLIER_ORDER_ID,A.ORDER_NO, W.LOCATION ,B.ITEM ,B.DIM_ITEM_ID,W.DIM_WAREHOUSE_ID,D.DIM_DATE_ID
                  FROM ITEM_SALES_ORDER A
                  JOIN DIM_ITEM B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  JOIN DIM_WAREHOUSE W
                  ON A.DIM_WAREHOUSE_ID=W.DIM_WAREHOUSE_ID
                  JOIN DIM_DATE D
                  ON A.ORDER_DATE=D.DATE_DESC
                  JOIN #DAYBOOK_STOCK_ITEM_NM C
                  ON B.ITEM=C.STOCK_ITEM_NAME
                  JOIN DIM_COMPANY E ON E.COMPANY_NAME=A.COMPANY_NAME
                 
UNION ALL
                  SELECT E.DIM_COMPANY_ID,A.RATE,A.JOURNAL_CREDIT_QUANTITY,(A.RATE*A.JOURNAL_CREDIT_QUANTITY) AS JOURNAL_CREDIT_VALUE,A.ORDER_DATE ASDATE_DESC, A.ITEM_SUPPLIER_ORDER_ID,A.ORDER_NO,W.LOCATION ,B.ITEM ,B.DIM_ITEM_ID,W.DIM_WAREHOUSE_ID,D.DIM_DATE_ID
                  FROM ITEM_JOURNAL_CREDIT A
                  JOIN DIM_ITEM B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  JOIN DIM_WAREHOUSE W
                  ON A.DIM_WAREHOUSE_ID=W.DIM_WAREHOUSE_ID
                  JOIN DIM_DATE D
                  ON A.ORDER_DATE=D.DATE_DESC
                  JOIN #DAYBOOK_STOCK_ITEM_NM C
                  ON B.ITEM=C.STOCK_ITEM_NAME
                  JOIN DIM_COMPANY E ON E.COMPANY_NAME=A.COMPANY_NAME
--where B.DIM_ITEM_ID=696
)
SELECT * INTO #CTE1 FROM CTE1;


--SELECT * FROM #CTE1

WITH CTE2 AS(
                  SELECT E.DIM_COMPANY_ID,A.RATE,A.PURCHASE_QUANTITY,(A.RATE*A.PURCHASE_QUANTITY) AS PURHASE_VALUE,A.ORDER_DATE ASDATE_DESC,A.ITEM_SUPPLIER_ORDER_ID,A.ORDER_NO, W.LOCATION ,B.ITEM ,B.DIM_ITEM_ID,W.DIM_WAREHOUSE_ID,D.DIM_DATE_ID
                  FROM ITEM_PURCHASE_ORDER A
                  JOIN DIM_ITEM B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  JOIN DIM_WAREHOUSE W
                  ON A.DIM_WAREHOUSE_ID=W.DIM_WAREHOUSE_ID
                  JOIN DIM_DATE D
                  ON A.ORDER_DATE=D.DATE_DESC
                  JOIN #DAYBOOK_STOCK_ITEM_NM C
                  ON B.ITEM=C.STOCK_ITEM_NAME
                  JOIN DIM_COMPANY E ON E.COMPANY_NAME=A.COMPANY_NAME
                  --where B.DIM_ITEM_ID=696
UNION ALL
                  SELECT E.DIM_COMPANY_ID,A.RATE,A.JOURNAL_DEBIT_QUANTITY,(A.RATE*A.JOURNAL_DEBIT_QUANTITY) AS JOURNAL_DEBIT_VALUE,A.ORDER_DATE ASDATE_DESC,A.ITEM_SUPPLIER_ORDER_ID,A.ORDER_NO, W.LOCATION ,B.ITEM ,B.DIM_ITEM_ID,W.DIM_WAREHOUSE_ID,D.DIM_DATE_ID
                  FROM ITEM_JOURNAL_DEBIT A
                  JOIN DIM_ITEM B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  JOIN DIM_WAREHOUSE W
                  ON A.DIM_WAREHOUSE_ID=W.DIM_WAREHOUSE_ID
                  JOIN DIM_DATE D
                  ON A.ORDER_DATE=D.DATE_DESC
                  JOIN #DAYBOOK_STOCK_ITEM_NM C
                  ON B.ITEM=C.STOCK_ITEM_NAME
                  JOIN DIM_COMPANY E ON E.COMPANY_NAME=A.COMPANY_NAME
                  --where B.DIM_ITEM_ID=696
)
SELECT * INTO #CTE2 FROM CTE2;

WITH
CTE3 AS(
                  SELECT DATE_DESC, LOCATION ,ITEM ,DIM_ITEM_ID,
                  DIM_WAREHOUSE_ID,
                  DIM_DATE_ID,DIM_COMPANY_ID
                  FROM  #CTE1
                  UNION ALL
                  SELECT DATE_DESC, LOCATION ,ITEM ,DIM_ITEM_ID,
                  DIM_WAREHOUSE_ID,
                  DIM_DATE_ID,DIM_COMPANY_ID
                  FROM #CTE2),
CTE4 AS(
                  SELECT ROW_NUMBER() OVER(ORDER BY
                  DIM_WAREHOUSE_ID ,
                  DIM_ITEM_ID,DIM_DATE_ID, DATE_DESC,DIM_COMPANY_ID)AS ROWID,DIM_DATE_ID,
                  DIM_WAREHOUSE_ID ,
                  DIM_ITEM_ID ,DATE_DESC,DIM_COMPANY_ID
                  FROM CTE3)

The dax code used to convert the sql code is as follows:
SUMMARIZE('Cumulative inventory total','Cumulative inventory total'[ITEM],'Cumulative inventory total'[ordeR_date],"total_purchase",sum('Cumulative inventory total'[quantity_debit]),"total_purchase_value",sum('Cumulative inventory total'[purchase_value_2]),"net_debit_credit_qty",sum('Cumulative inventory total'[quantity_debit_credit]))


Explanation of the quantity_debit: if('Cumulative inventory total'[is_deemed_positive]="yes",'Cumulative inventory total'[quantity],0)
Purchase_value_2: IF('Cumulative inventory total'[is_deemed_positive]="yes",'Cumulative inventory total'[QUANTITY_APPENDED]*'Cumulative inventory total'[RATE_APPENDED],0)
Net_debit_credit_qty: SUMMARIZE('Cumulative inventory total','Cumulative inventory total'[ITEM],'Cumulative inventory total'[ordeR_date],"total_purchase",sum('Cumulative inventory total'[quantity_debit]),"total_purchase_value",sum('Cumulative inventory total'[purchase_value_2]),"net_debit_credit_qty",sum('Cumulative inventory total'[quantity_debit_credit]))

Removal of the perpetual weighted average inventory calculations using complex while do loop across all the stock items as given. For the audit purposes the end clients expected the inventory valuations on a daily basis to catch any data entry errors.
The dax code uses running sum across all date ranges for the stock items to capture the historical stock valuations.

CTE5 AS(
                  SELECT MIN(ROWID)AS MINROW,DIM_DATE_ID,
                  DIM_WAREHOUSE_ID ,
                  DIM_ITEM_ID ,DATE_DESC,DIM_COMPANY_ID
                  FROM CTE4
                  GROUP BY DIM_DATE_ID,
                  DIM_WAREHOUSE_ID ,
                  DIM_ITEM_ID,DATE_DESC,DIM_COMPANY_ID)
           
INSERT INTO [MOVING_AVERAGE_RATE]( DIM_DATE_ID,DIM_ITEM_ID,DATE_DESC,DIM_COMPANY_ID)
                  SELECT  DIM_DATE_ID,DIM_ITEM_ID,DATE_DESC,DIM_COMPANY_ID
                  FROM CTE5
                  ORDER BY  DIM_ITEM_ID,DIM_DATE_ID, DATE_DESC,DIM_COMPANY_ID;

WITH CTE_PURCHASE AS (
                  SELECT SUM(PURCHASE_QUANTITY) AS PURCHASE_QUANTITY,SUM(PURHASE_VALUE) AS PURCHASE_VALUE,
                  DIM_ITEM_ID,DIM_DATE_ID,DIM_COMPANY_ID,DIM_WAREHOUSE_ID
                  FROM #CTE2
                  GROUP BY DIM_ITEM_ID,DIM_DATE_ID,DIM_COMPANY_ID,DIM_WAREHOUSE_ID)
                 
UPDATE MOVING_AVERAGE_RATE
                  SET PURCHASE_QUANTITY= ISNULL(B.PURCHASE_QUANTITY,0),
                  PURCHASE_VALUE=ISNULL(B.PURCHASE_VALUE,0)
                  FROM MOVING_AVERAGE_RATE A
                  LEFT OUTER JOIN CTE_PURCHASE B
                  ON
                  --A.DIM_WAREHOUSE_ID=B.DIM_WAREHOUSE_ID
                  --AND
                  A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  AND A.DIM_DATE_ID=B.DIM_DATE_ID
                  AND A.DIM_COMPANY_ID=B.DIM_COMPANY_ID;
           
WITH CTE_SALES AS (
                  SELECT SUM(SALE_QUANTITY) AS SALE_QUANTITY,SUM(SALE_VALUE) AS SALE_VALUE,
                  DIM_ITEM_ID,DIM_DATE_ID,DIM_COMPANY_ID,DIM_WAREHOUSE_ID
                  FROM #CTE1
                  GROUP BY DIM_ITEM_ID,DIM_DATE_ID,DIM_COMPANY_ID,DIM_WAREHOUSE_ID)

UPDATE MOVING_AVERAGE_RATE
                  SET SALE_QUANTITY= ISNULL(B.SALE_QUANTITY,0),
                  SALE_VALUE=ISNULL(B.SALE_VALUE,0)
                  FROM MOVING_AVERAGE_RATE A
                  LEFT OUTER JOIN CTE_SALES B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  AND A.DIM_DATE_ID=B.DIM_DATE_ID
                  AND A.DIM_COMPANY_ID=B.DIM_COMPANY_ID;
                 
WITH CTE1
AS(SELECT ROW_NUMBER() OVER(OrdeR BY ltrim(rtrim(replace(REPLACE(item_name,char(13),''),char(10),'')))) AS ROW_NUMBER2,
ltrim(rtrim(replace(REPLACE(item_name,char(13),''),char(10),'')))  AS item_name,
CAST(QTY AS FLOAT)AS QTY,CAST(VALUE AS FLOAT) AS VALUE,CAST(RATE AS  FLOAT) AS RATE FROM dbo.OPENING_STOCK
WHERE VALUE IS NOT NULL
GROUP BY ltrim(rtrim(replace(REPLACE(item_name,char(13),''),char(10),'')))
,CAST(VALUE AS FLOAT),CAST(QTY AS FLOAT),CAST(RATE AS  FLOAT)

),
CTE2
AS(SELECT MIN(ROW_NUMBER2) AS MIN_ROW_ID,
item_name
  FROM CTE1
GROUP BY item_name),
CTE3
AS(SELECT  A.* FROM CTE1 A
JOIN CTE2 B ON
A.ROW_NUMBER2=B.MIN_ROW_ID
--WHERE A.QTY IS NOT NULL
)
           
UPDATE MOVING_AVERAGE_RATE
                  SET OPENING_STOCK=ISNULL(CAST(CTE3.QTY AS FLOAT),0),
OPENING_STOCK_VALUE=ISNULL(CAST(CTE3.VALUE AS FLOAT),0)
           
      ,          

MOVING_AVERAGE_RATE=CASE WHEN (ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(CAST(CTE3.QTY AS FLOAT),0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN 0
                                            ELSE  ((ISNULL(CAST(CTE3.VALUE AS FLOAT),0)+(PURCHASE_VALUE))/(ISNULL(CAST(CTE3.QTY ASFLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)))
END

/*CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=0 THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=0 AND CAST(PURCHASE_QUANTITY AS FLOAT)>0 THEN CAST(PURCHASE_VALUE AS FLOAT)/CAST(PURCHASE_QUANTITY AS FLOAT)
                                            WHEN ISNULL(OPENING_STOCK,0) >0 AND  PURCHASE_QUANTITY >0 THEN ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))

ELSE (ISNULL(OPENING_STOCK_VALUE,0))/CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)


                                           END*/



,


                  --COST_OF_GOODS_SOLD = CASE WHEN PURCHASE_QUANTITY=0 THEN 0 ELSE SALE_QUANTITY*(PURCHASE_VALUE/PURCHASE_QUANTITY) END,
      /*    COST_OF_GOODS_SOLD=(CASE
--WHEN (OPENING_STOCK+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=0 THEN 0
                                          WHEN CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)=0 AND CAST(PURCHASE_QUANTITY AS FLOAT)>0 THEN CAST(PURCHASE_VALUE AS FLOAT)/CAST(PURCHASE_QUANTITY AS FLOAT)
                                            WHEN CAST(ISNULL(OPENING_STOCK,0) AS FLOAT) >0 AND  CAST(PURCHASE_QUANTITY AS FLOAT) >0 THEN ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))

ELSE (ISNULL(OPENING_STOCK_VALUE,0))/CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)


                                           END)*SALE_QUANTITY,*/
                 
CLOSING_STOCK_VALUE= ((ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT)))*

(CASE WHEN (ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(CAST(CTE3.QTY AS FLOAT),0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN 0
                                            ELSE  ((ISNULL(CAST(CTE3.VALUE AS FLOAT),0)+(PURCHASE_VALUE))/(ISNULL(CAST(CTE3.QTY ASFLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)))
END),




/*(CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=0 THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=0 AND CAST(PURCHASE_QUANTITY AS FLOAT)>0 THEN CAST(PURCHASE_VALUE AS FLOAT)/CAST(PURCHASE_QUANTITY AS FLOAT)
                                            WHEN ISNULL(OPENING_STOCK,0) >0 AND  PURCHASE_QUANTITY >0 THEN ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))

ELSE (ISNULL(OPENING_STOCK_VALUE,0))/CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)


                                           END)*(CAST(ISNULL(OPENING_STOCK,0) AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT)),*/
CLOSING_STOCK=(ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT)),

COST_OF_GOODS_SOLD = ISNULL(CAST(CTE3.VALUE AS FLOAT),0)+CAST(PURCHASE_VALUE AS FLOAT)-((ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY ASFLOAT)-CAST(SALE_QUANTITY AS FLOAT)))*

(CASE WHEN (ISNULL(CAST(CTE3.QTY AS FLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(CAST(CTE3.QTY AS FLOAT),0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN 0
                                            ELSE  ((ISNULL(CAST(CTE3.VALUE AS FLOAT),0))+(PURCHASE_VALUE))/(ISNULL(CAST(CTE3.QTY ASFLOAT),0)+CAST(PURCHASE_QUANTITY AS FLOAT))
END)
                  FROM MOVING_AVERAGE_RATE A
                  JOIN DIM_ITEM B
                  ON A.DIM_ITEM_ID=B.DIM_ITEM_ID
                  LEFT OUTER JOIN CTE3 ON CTE3.ITEM_NAME=B.ITEM
                  WHERE MOVING_AVERAGE_RATE_ID IN (SELECT MIN(MOVING_AVERAGE_RATE_ID) FROM MOVING_AVERAGE_RATE A
GROUP BY DIM_ITEM_ID);

DELETE FROM MOVING_AVERAGE_RATE WHERE CLOSING_STOCK<0;

DECLARE @COUNT INT;
DECLARE @COUNT_ITEM_START INT;
DECLARE @COUNT_ITEM INT;
DECLARE @CHECKNO INT;
DECLARE @MOVING_AVERAGE_RATE DECIMAL(15,3);
DECLARE @CLOSING_STOCK DECIMAL(15,3);
DECLARE @CLOSING_STOCK_VALUE DECIMAL(15,3);

--DECLARE @MIN_DIM_WAREHOUSE_ID INT;
--SET @MIN_DIM_WAREHOUSE_ID=(SELECT MIN(DIM_WAREHOUSE_ID) FROM OPENING_STOCK);
--DECLARE @MAX_DIM_WAREHOUSE_ID INT;
--SET @MAX_DIM_WAREHOUSE_ID=(SELECT MAX(DIM_WAREHOUSE_ID) FROM OPENING_STOCK);
SET @COUNT_ITEM_START=1;
--WHILE(@MIN_DIM_WAREHOUSE_ID<=@MAX_DIM_WAREHOUSE_ID)
--BEGIN

SET @COUNT_ITEM_START=1;
SET @COUNT_ITEM=(SELECT COUNT(DIM_ITEM_ID) FROM DIM_ITEM);
WHILE @COUNT_ITEM_START<=@COUNT_ITEM
BEGIN
SET @COUNT=(SELECT MAX(MOVING_AVERAGE_RATE_ID) FROM MOVING_AVERAGE_RATE WHERE DIM_ITEM_ID=@COUNT_ITEM_START)
SET @CHECKNO=(SELECT MIN(MOVING_AVERAGE_RATE_ID)+FROM MOVING_AVERAGE_RATE WHERE DIM_ITEM_ID=@COUNT_ITEM_START) ;
WHILE(@CHECKNO<=@COUNT)
BEGIN
UPDATE MOVING_AVERAGE_RATE
                  SET OPENING_STOCK=S.CLOSING_STOCK,
                  OPENING_STOCK_VALUE=S.CLOSING_STOCK_VALUE,
@MOVING_AVERAGE_RATE=S.MOVING_AVERAGE_RATE,
            --    ,COST_OF_GOODS_SOLD=S.MOVING_AVERAGE_RATE*A.SALE_QUANTITY,
@CLOSING_STOCK=ISNULL(S.CLOSING_STOCK,0),
@CLOSING_STOCK_VALUE=ISNULL(S.CLOSING_STOCK_VALUE,0)
                  FROM MOVING_AVERAGE_RATE A
                  JOIN (SELECT A.MOVING_AVERAGE_RATE_ID,B.CLOSING_STOCK,B.CLOSING_STOCK_VALUE,B.MOVING_AVERAGE_RATE
                              FROM MOVING_AVERAGE_RATE A
                              JOIN MOVING_AVERAGE_RATE B
                              ON A.MOVING_AVERAGE_RATE_ID=B.MOVING_AVERAGE_RATE_ID+1
                                    WHERE A.MOVING_AVERAGE_RATE_ID IN (SELECT MIN(MOVING_AVERAGE_RATE_ID) FROM MOVING_AVERAGE_RATE
                                                                              WHERE OPENING_STOCK IS NULL
AND DIM_ITEM_ID=@COUNT_ITEM_START
                                                                              GROUP BY DIM_ITEM_ID,DIM_DATE_ID)
                        )S
                  ON S.MOVING_AVERAGE_RATE_ID=A.MOVING_AVERAGE_RATE_ID
                  JOIN DIM_ITEM D
                  ON A.DIM_ITEM_ID=D.DIM_ITEM_ID
AND A.DIM_ITEM_ID=@COUNT_ITEM_START
                  AND D.ITEM IN (SELECT STOCK_ITEM_NAME FROM #DAYBOOK_STOCK_ITEM_NM)
                  --AND A.DIM_WAREHOUSE_ID=@MIN_DIM_WAREHOUSE_ID
--AND (ISNULL(A.OPENING_STOCK,0)+CAST(A.PURCHASE_QUANTITY AS FLOAT)-CAST(A.SALE_QUANTITY AS FLOAT))>=0

UPDATE MOVING_AVERAGE_RATE
SET
CLOSING_STOCK=OPENING_STOCK+PURCHASE_QUANTITY-SALE_QUANTITY,
MOVING_AVERAGE_RATE=CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN @MOVING_AVERAGE_RATE
WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))<THEN @MOVING_AVERAGE_RATE*(-1)
                                            ELSE  ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(OPENING_STOCK AS FLOAT)+CAST(PURCHASE_QUANTITYAS FLOAT)))
END,
/*COST_OF_GOODS_SOLD=

/*(CASE WHEN (ISNULL(@CLOSING_STOCK,0)+CAST(ISNULL(PURCHASE_QUANTITY,0) AS FLOAT)-CAST(ISNULL(SALE_QUANTITY,0) AS FLOAT))=0 THEN 0
                                          WHEN ISNULL(@CLOSING_STOCK,0)=0 AND CAST(ISNULL(PURCHASE_QUANTITY,0) AS FLOAT)>0 THEN CAST(ISNULL(PURCHASE_VALUE,0) AS FLOAT)/CAST(PURCHASE_QUANTITY AS FLOAT)
                                            WHEN ISNULL(@CLOSING_STOCK,0)>0 AND  ISNULL(PURCHASE_QUANTITY,0) >0 THEN ((ISNULL(@CLOSING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(ISNULL(@CLOSING_STOCK,0) AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))

WHEN ISNULL(@CLOSING_STOCK,0) >0 AND  ISNULL(PURCHASE_QUANTITY,0) =0  THEN (ISNULL(@CLOSING_STOCK_VALUE,0))/CAST(ISNULL(@CLOSING_STOCK,0) AS FLOAT)
ELSE @MOVING_AVERAGE_RATE END)*/

(CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=0 THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=0 AND CAST(PURCHASE_QUANTITY AS FLOAT)=0 THEN @MOVING_AVERAGE_RATE
                                            ELSE  ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(OPENING_STOCK AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))
END)*A.SALE_QUANTITY,*/

                                               
                  --CLOSING_STOCK_VALUE= OPENING_STOCK_VALUE+PURCHASE_VALUE-COST_OF_GOODS_SOLD
CLOSING_STOCK_VALUE=CASE WHEN (OPENING_STOCK+PURCHASE_QUANTITY-SALE_QUANTITY=0) THEN 0
ELSE

(OPENING_STOCK_VALUE+PURCHASE_VALUE-
((CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN @MOVING_AVERAGE_RATE
WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))<THEN @MOVING_AVERAGE_RATE*(-1)
                                            ELSE  ((ISNULL(OPENING_STOCK_VALUE,0)+(PURCHASE_VALUE))/(CAST(OPENING_STOCK AS FLOAT)+CAST(PURCHASE_QUANTITYAS FLOAT)))
END)*A.SALE_QUANTITY)) END ,


COST_OF_GOODS_SOLD=
(CASE WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))=THEN 0
                                          WHEN ISNULL(OPENING_STOCK,0)=AND CAST(PURCHASE_QUANTITY AS FLOAT)=THEN @MOVING_AVERAGE_RATE
WHEN (ISNULL(OPENING_STOCK,0)+CAST(PURCHASE_QUANTITY AS FLOAT)-CAST(SALE_QUANTITY AS FLOAT))<THEN @MOVING_AVERAGE_RATE*(-1)                                     
 ELSE 
((ISNULL(OPENING_STOCK_VALUE,0)+PURCHASE_VALUE))/(CAST(OPENING_STOCK AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT))
END)*A.SALE_QUANTITY



/*ISNULL(@CLOSING_STOCK_VALUE,0)+PURCHASE_VALUE-(CASE WHEN (ISNULL(@CLOSING_STOCK,0)+CAST(ISNULL(PURCHASE_QUANTITY,0) AS FLOAT)-CAST(ISNULL(SALE_QUANTITY,0) AS FLOAT))=0 THEN 0
                                          WHEN ISNULL(@CLOSING_STOCK,0)=0 AND CAST(PURCHASE_QUANTITY AS FLOAT)>0 THEN CAST(PURCHASE_VALUE AS FLOAT)/CAST(PURCHASE_QUANTITY AS FLOAT)
                                            WHEN @CLOSING_STOCK >0 AND  PURCHASE_QUANTITY >0 THEN ((@CLOSING_STOCK_VALUE+(PURCHASE_VALUE))/(CAST(@CLOSING_STOCK AS FLOAT)+CAST(PURCHASE_QUANTITY AS FLOAT)))

WHEN @CLOSING_STOCK >0 AND  PURCHASE_QUANTITY =0  THEN (@CLOSING_STOCK_VALUE)/CAST(@CLOSING_STOCK AS FLOAT)
ELSE @MOVING_AVERAGE_RATE  END)*A.SALE_QUANTITY*/


                  FROM MOVING_AVERAGE_RATE A
                  JOIN DIM_ITEM D
                  ON A.DIM_ITEM_ID=D.DIM_ITEM_ID
AND A.DIM_ITEM_ID=@COUNT_ITEM_START
--AND A.DIM_WAREHOUSE_ID=@MIN_DIM_WAREHOUSE_ID
                  WHERE A.MOVING_AVERAGE_RATE_ID IN (SELECT MAX(MOVING_AVERAGE_RATE_ID) FROM MOVING_AVERAGE_RATE
                                                                              WHERE OPENING_STOCK IS NOT NULL
AND DIM_ITEM_ID=@COUNT_ITEM_START
--AND (ISNULL(A.OPENING_STOCK,0)+CAST(A.PURCHASE_QUANTITY AS FLOAT)-CAST(A.SALE_QUANTITY AS FLOAT))>=0
                                                                              GROUP BY DIM_ITEM_ID)
                  AND D.ITEM IN (SELECT STOCK_ITEM_NAME FROM #DAYBOOK_STOCK_ITEM_NM);
SET @CLOSING_STOCK=(SELECT (OPENING_STOCK+PURCHASE_QUANTITY-SALE_QUANTITY) FROM MOVING_AVERAGE_RATE
WHERE DIM_ITEM_ID=@COUNT_ITEM_START
AND MOVING_AVERAGE_RATE_ID=@CHECKNO);
IF @CLOSING_STOCK<BREAK;
SET @CHECKNO=@CHECKNO+1;
END;
SET @COUNT_ITEM_START=@COUNT_ITEM_START+1;
PRINT 'The following item id has been loaded '+CAST(@COUNT_ITEM_START AS VARCHAR);
END;
--SET @MIN_DIM_WAREHOUSE_ID=@MIN_DIM_WAREHOUSE_ID+1;
--PRINT 'The following warehouse id has been loaded '+CAST(@MIN_DIM_WAREHOUSE_ID AS VARCHAR);
--END;
END;


The following is the dax code for the same:
Cumulative of purchase quantity:
cumulative_purchase_no = CALCULATE(SUM('table2'[total_purchase]),filter(allexcept('table2','table2'[ITEM]),'table2'[ordeR_date]<=max('table2'[ordeR_date])))

cumulative of purchase value:
cumulative_purchase value = CALCULATE(SUM('table2'[total_purchase_value]),filter(allexcept('table2','table2'[ITEM]),'table2'[ordeR_date]<=max('table2'[ordeR_date])))

cumulative_purchase value = CALCULATE(SUM('table2'[total_purchase_value]),filter(allexcept('table2','table2'[ITEM]),'table2'[ordeR_date]<=max('table2'[ordeR_date])))

cumulative_purchase_qty2 = 'table2'[cumulative_purchase_no]+'table2'[opening_stock qty]

net_closing_qty =
SUM('table2'[net_debit_credit_qty]) + SUM('table2'[opening_stock qty])

opening_stock qty = LOOKUPVALUE(openinG_stock[DSPCLQTY],openinG_stock[DSPDISPNAME],'table2'[ITEM])

opening_stock rate = LOOKUPVALUE(openinG_stock[DSPCLRATE],openinG_stock[DSPDISPNAME],'table2'[ITEM])


weighted_average =
DIVIDE(
    [cumulative_purchase value2],
    SUM('table2'[cumulative_purchase_qty2])
)

Final result:

Results of the code:
·         Attached are the final results for the process:
o   Reduced execution times
o   Reduced times for the weighted average calculations
o   Better error handling and logging capabilities
o   Reduced deployment requirements


Prepared by :
Ankur
Email: apoorv@mnnbi.com
website: www.mnnbi.com
Phone: +91-8802466356

(For additional details on how PowerBI could be used for better reporting, contact above)

















Initial and Incremental data Load Template by M&N Business Intelligence-SAP Data Services

  INCREMENTAL LOAD/CREATION OF DIMENSION TABLE LOGIC AT SAP DATA SERVICES END CLIENT In this particular document we shall be looking at the ...