Forum Discussion
Helper VPivot Table in Excel to PowerBI
Helper Vlbendlin did you had a look at my data yet? https://docs.google.com/spreadsheets/d/1-FfCgSXpK5fNyH3AWyvwfUwBXT9WQqBb/edit?usp=sharing&ouid=104129043494164133703&rtpof=true&sd=true
Super Useryes. I am struggling with point 3. Is SAP Data your raw data?
- Yrstruly2021
SAP Data will be the transactional data that is joined with dimensional data/tables. The SAP source data is then pivoted, as seen in 1st image above.
My challenge is, building the pivoted table in PowerBI from the SQL data source(extract from SQL), see:https://docs.google.com/spreadsheets/d/12RjkTzq1elO3bRzvu4LZ8DxHypZlzIsE/edit?usp=sharing&ouid=104129043494164133703&rtpof=true&sd=true
The Matrix/Pivot feature in Powerbi will work. Now it is building the formulas like, e.g "=-SUMIF('SAP Source'!C[-10];'Cash Flow'!RC[-4];'SAP Source'!C[-3])+334776"
into PowerBI. So all of the referencing is shifting from Excel sheets to SQL table(s).
The data link(https://docs.google.com/spreadsheets/d/1-FfCgSXpK5fNyH3AWyvwfUwBXT9WQqBb/edit?usp=sharing&ouid=104129043494164133703&rtpof=true&sd=true ) contains(SAP + Dimension tables) the tab Cash Flow where you can investigate how the values in Column G are compiled. The primary key contains duplicates, these can be filtered.
Data referential integrity is Many - 1. There can be multiple Account_Number linked to 1
ZTBR_TransactionCode.
Please advise on optimisation methods.
- lbendlin
The raw data is usable. Your formulas however are not. Please describe the business processes rather than the Excel formulas. Power BI does not have an easy way to reference "same row, four columns to the left" and similar.
- Yrstruly2021
Would you be able to replicate the PowerBI report with this data, https://drive.google.com/file/d/17aOmG-Ynx-6U05wNNrHXJG7iywgCMiuy/view?usp=sharing ?
(1) The fact table is from SAP.
(2) Dimension Tables are named "B Masked". Please note that none of the tables had primary keys, I had to bring them in myself. In the Bracs Mapping tab of this file, there is a column GCoA which is the Account Number. Then there is BRACS which is the Bracs(source) Account number. these numbers match the same columns in the Mapping tab of the Cash Flow Pivoted file.
(3) The Source Data is a combination and shorter version of the Mapping tab. Column Account match GCoA and BRACS Account match
matches BRACS in the Bracs mapping tab above of the B Masked file.
I noticed some similarities in the different tabs/tables so I denormalised some tabs/tables. e.g. Region Mapping and Entity Mapping in the same file I made as 1 table.
Joining on not unique keys is of concern.
(4) Cash Flow file, tab Sap Source is a pivoted table of the Source Data.
(5) The Cash Flow tab is the final reporting structure that must be replicated in power. Column "G" in this tab is the "Function" column in the Pivoted tab.
I do left joins to the Fact table and some inner join. Please advise.
The Function column that makes up the bulk of giving meaning to the figures, most of the detail of this column is left out when I do a join as seen in Imeta_Bracs_Roll_Up joined to Fact file with SQL code.
I brought in the Mapping table to connect the roll-up table to it as seen in Imeta_Bracs_Roll_Up joined to Bridge, this query takes. very long to run. Also, the joins take a lot of memory and some ETL/SSIS tasks may fail. What would you be bettering in this instance?
PBIX https://drive.google.com/file/d/1_XyrEdRUsQImyuDhapkA5fqcpXRGUjRV/view?usp=sharing