Package: splitGraph 0.4.0

What this vignette is for | Build a split_spec to work with | Adapter 1 — base R: leave-one-group-out folds | Honoring block variables | Honoring the stratum annotation | Adapter 2 — rsample::group_vfold_cv() | Adapter 3 — rsample::rolling_origin() | Two things to check before trusting an ordered adapter | Going across language boundaries via JSON | Letting the spec pick the adapter | When you need a custom adapter

The data | Build the graph | What validation says | How much would a naive split leak? | Derive constraints | Composite grouping and over-merging | The split_spec | Working with a subset | Handing off

Derive and serialize in R | What is actually on disk | Read in Python | Versions are part of the contract | Verify the round-trip | Drive scikit-learn | Why this matters

Why not just call make_split_plan(group = , batch = ) directly? | When does composite-strict over-merge? | How do thresholds interact with transitive closure? | What does the stratum column mean, and does splitGraph stratify? | Schema versioning policy, in one place | Can I build a graph straight from a SummarizedExperiment? | Which errors can I catch programmatically? | How large a cohort can splitGraph handle?

Cluster-style relations: site, region, platform, assay | Watch for a composite that collapses | Pairwise relation: genetic relatedness | Pairwise relation: spatial proximity | Deriving on a subset is leakage-safe | Combining a pairwise relation with a direct one | Thresholds are inputs, not modeling

1. A metadata table | 2. Build and validate | 3. Choose a constraint mode | 4. Emit the split_spec | 5. See which leakage paths the choice closes | 6. Hand off | Where to go next

Why splitGraph exists | A realistic toy dataset | Fast path: graph_from_metadata() | Ingest metadata and build typed nodes and edges | Assemble the dependency graph | Visualize the typed structure | Reshaping a graph you already have | Exporting for other tools | Validate before you split | When you really do need to relax a check | Query the graph to inspect hidden structure | Derive split constraints from the graph | Batch constraints | Time constraints | Composite constraints | Cluster-style relations: site, region, platform, assay | Site | Region | Platform and assay | Pairwise relations: relatedness and spatial proximity | Time ordering can come from precedence edges alone | Translate the constraint into a split specification | Summarize the leakage picture in one object | Downstream handoff | Persisting the handoff | Case studies | Case study 1: repeated subjects in a longitudinal cohort | Case study 2: a subject reused across studies | Case study 3: partially observed technical metadata | Case study 4: choosing a defensible split strategy | When splitGraph is useful | What splitGraph is not for | Takeaway