Skip to contents

Set partitioning for a DuckLake table

Source: R/discovery.R
db_set_partitioning.Rd

Configures partition keys for a DuckLake table. When partitioning is set, new data written to the table will be split into separate files based on the partition key values.

Partitioning enables:

  • Efficient query pruning (only read relevant partitions)

  • Potential folder-based access control at partition level

  • Better data organization for time-series data

Note: Existing data is not reorganized - only new inserts are partitioned.

Usage

db_set_partitioning(schema = "main", table, partition_by)

Arguments

schema

Schema name

table

Table name

partition_by

Character vector of column names or expressions to partition by. Use NULL to remove partitioning.

Value

Invisibly returns TRUE on success

Examples

if (FALSE) { # \dontrun{
db_lake_connect(...)

# Partition by year and month columns
db_set_partitioning("trade", "imports", c("year", "month"))

# Partition using date functions
db_set_partitioning("trade", "imports", c("year(date)", "month(date)"))

# Remove partitioning (new data won't be partitioned)
db_set_partitioning("trade", "imports", NULL)
} # }