Split-methods: Split an item pool into partitions

In TestDesign: Optimal Test Design Approach to Fixed and Adaptive Test Construction

Split an item pool into partitions

Description

Split is a function to split a pool into multiple parallel tests or pools. When constructing parallel tests, each test is constructed to satisfy all constraints. When constructing parallel pools, each pool is constructed so that it contains a test that satisfies all constraints.

Usage

Split(
  config,
  constraints,
  n_partition,
  partition_type,
  partition_size_range = NULL,
  force_solver = FALSE
)

## S4 method for signature 'config_Static'
Split(
  config,
  constraints,
  n_partition,
  partition_type,
  partition_size_range = NULL,
  force_solver = FALSE
)

Arguments

config	a config_Static object. Use createStaticTestConfig for this.
constraints	a constraints object representing test specifications. Use loadConstraints for this.
n_partition	the number of partitions to create.
partition_type	test to create tests, or pool to create pools.
partition_size_range	(optional) two integer values for the desired range for the size of a partition. Has no effect when partition_type is test. For discrete item pools, the default partition size is (pool size / number of partitions). For set-based item pools, the default partition size is (pool size / number of partitions) +/- smallest set size.
force_solver	if TRUE, do not check whether the solver is one of recommended solvers for complex problems (set-based assembly, partitioning). (default = FALSE)

Value

partition returns an output_Split object containing item/set indices of created tests/pools.

Examples

## Not run: 
config <- createStaticTestConfig(MIP = list(solver = "RSYMPHONY"))
constraints <- constraints_science[1:10]

solution <- Split(config, constraints, n_partition = 4, partition_type = "test"))
plot(solution)
solution <- Split(config, constraints, n_partition = 4, partition_type = "pool"))
plot(solution)

## End(Not run)

TestDesign documentation built on Feb. 16, 2023, 7:19 p.m.