Frame the month-end question and inspect five sources

Import the SAP scenario and distinguish the role of each of its five source tables.

Question for this chapter

Which tables must be connected to explain revenue?

Why this matters now

Starting with a total from one table removes its customer and material context. First identify the role and join keys of the five sources so that you can later explain what each result means.

Try it

sap.zip Download(1.6 MB)

Select Collections in the left sidebar, then choose More (⋯) → Import in the upper-right. Upload the ZIP and open the SAP Data Collection.

Korean Portal screen with Import selected from a collection's More menu — Collection import menu in the Korean Portal used for the verified run

Open SAP Sales Document Header (VBAK) in the collection and select the Data tab.

Data tab of the SAP sales-document header dataset in the Korean Portal — VBAK contains the header information for one order.

Visible alias	SAP abbreviation	Role	Join key
SAP Sales Document Header	VBAK	Order header with date and customer	Order number
SAP Sales Document Item	VBAP	Material, quantity, and amount by order line	Order number, material number
SAP Material Master	MARA	Material type and base attributes	Material number
SAP Material Description	MAKT	Language-specific material name	Material number
SAP Customer Master	KNA1	Customer name and region	Customer number

Deep dive — complete imported asset inventory

Asset type	Count	Contents
Collection	1	SAP Data Collection (`sap`)
Datasets	22	Five source datasets and 17 derived datasets rebuilt by pipelines
Code	19	Sales, material, and customer preparation; summaries; order review; ontology materialization
Pipelines	14	Execution paths from source preparation through order review and ontology materialization
Ontology	1 model / 10 entities / 10 relations	Customer → order → line → material plus date and region hierarchies
Tool and Agent	1 each	`ontology_graph_query`, `sap_assistant`

The ZIP uploads files only to the five source datasets. The pipelines rebuild the other 17 datasets from those sources, so empty derived datasets immediately after import are expected.

Success looks like this

The SAP Data Collection contains the asset groups shown above, and the five source datasets contain data. Only those five source files are in the ZIP; the pipelines recreate every derived dataset.

Interpret the result

VBAK joins to VBAP by order number, VBAP joins to MARA and MAKT by material number, and VBAK joins to KNA1 by customer number. This is what lets one line amount acquire customer and material meaning.

Next decision

The source codes and labels are not yet ready for analysis. Next, prepare the sales, material, and customer domains separately.