H25 Configure a local DuckLake

Use your current course session. Preparation and course runner · R script ↓

These exercises need DuckLake. Enable the lake option and prepare the listed packages as described in the script. For a local alternative without the DuckLake extension, see V25.

H25 Configure a local DuckLake

Goal: Separate metadata from data-file storage.

Prerequisites: lake=TRUE; dataraft.lake.

DuckDB is the execution engine. DuckLake manages tables through a metadata catalog and data files. The local path shorthand defaults to DuckDB without an explicit backend. This creates a new DuckLake; it does not migrate an existing DuckDB directory.

training_step(
  "H25",
  {
    lake_config <- dataraft.lake::dr_lake_config(
      path = file.path(training_root, "ducklake"),
      backend = "ducklake",
      install_extensions = training_options$install_extensions
    )
    print(lake_config)
  },
  packages = c("dataraft.lake"),
  enabled = training_options$lake,
  optional = TRUE
)

Expected result: Configuration only, with no open connection.

Task: Where will the data live?