Resolving Orphan Batches and Ambiguous Batches During Signals Data Factory Migration

• During pre-migration validation, the project displays a count of Orphan Batches and/or Ambiguous Batches that does not decrease even after uploading correction files.
• The same Batch ID appears under both orphan batch and ambiguous batch warning lists.
• A circular blocker occurs: following the standard recommendation to "clear orphan batches first before fixing ambiguous batches" stalls because the batch relationship cannot be validated without a compound mapping.
• In projects with parent/child Measurement Types, child datasets contain only Batch IDs without a Compound ID column, while parent datasets contain Compound IDs with missing or duplicate associations.

• SDF 2.0 Relationship Architecture: Unlike legacy pipelines that may infer compound relationships across joined tables, SDF 2.0 requires every measurement dataset to possess a direct compound association. If a child MT dataset does not define a Compound ID column, its batches are flagged as orphans.
• The "One Parent per Batch" Rule: Across an entire SDF project, a single Batch ID (e.g., B001) can have exactly one parent Compound ID (e.g., C001). If B001 is mapped to C001 in one row or file, but to C002 in another, it triggers an Ambiguous Batch warning.
• Empty Compound ID Mapping: If a dataset has the Compound ID column mapped, but certain rows have empty/blank cells, Data Factory evaluates the parent as <EMPTY>-B001. If another file or an orphan correction file maps C001-B001, the system registers two conflicting parents (<EMPTY> and C001), creating an ambiguous batch conflict.
• Replicates Are Allowed: Multiple rows having identical compound and batch combinations (e.g., replicates of C01-B01) are fully supported within or across datasets. The conflict only arises when the parent compound changes for the same batch.

• Identify Affected Datasets & Rows: Revvity Signals Support and Engineering can run diagnostic scripts to extract detailed CSV reports listing every affected dataset, internal dataset identifier (selectedDataset), row index, Batch ID, and Compound ID.
• Update Child Datasets: Add explicit Compound ID and Batch ID columns directly to child MT source files.
• Standardize Parent Datasets: Eliminate empty, missing, or contradictory Compound ID values in parent source files.
• Perform a Full Reload (LoadAll): Re-execute a full load so legacy SDF synchronizes and registers the updated source relationships.

Q: Can a single batch appear multiple times in a dataset?
Yes. Multiple replicate measurements for the same compound and batch combination (e.g., multiple rows of C001 / B001) are fully valid and supported.

Q: Why did uploading an "Orphan Batch Update" file create ambiguous batches?
If your source dataset already had the Compound ID column mapped but left blank (<EMPTY>), the system considers <EMPTY> to be Parent 1. When your correction file introduced C001 as Parent 2, a conflict was created. You must populate or unmap the blank cell in the source file itself.

Q: Do child datasets really need Compound IDs if the parent already defines them?
Yes. In SDF 2.0 architecture, measurements are verified directly against compound definitions. Child measurement datasets must explicitly provide their parent Compound ID and Batch ID.

Q: What is selectedDataset or Dataset ID in the diagnostic report?
In Signals Inventa / Data Factory, each published table or spreadsheet is assigned an internal UUID/ID known as selectedDataset. Support diagnostic reports list this ID alongside the dataset display name so you can pinpoint the exact file in your project configuration.