rand_partitions: Generate uniform random partitions of Q having N parts.
In rpartitions: Code for integer partitioning

Description Usage Arguments Value Note Examples

Generate uniform random partitions of Q having N parts.

1
2
3

  rand_partitions(Q, N, sample_size, method = "best",
    D = hash(), zeros = FALSE, use_c = TRUE,
    use_hash = FALSE)

`Q`	Total sum across parts
`N`	Number of parts to sum over
`sample_size`	number of random partitions to generate
`method`	: method to use for generating the partition, options include: 'bottom_up', 'top_down', 'divide_and_conquer', 'multiplicity', and 'best'. Defaults to 'best'
`D`	a dictionary for the number of partitions of Q having N or less parts (or N or less as the largest part), i.e. P(Q, Q + N). Defaults to a blank dictionary.
`zeros`	boolean if True partitions can have zero values, if False partitions have only positive values, defaults to False
`use_c`	boolean if TRUE then compiled c code is used, defaults to TRUE
`use_hash`	boolean, if TRUE then a hash table is used, defaults to FALSE

A matrix where each column is a random partition

method 'best' attempts to use the values of Q and N to infer what the fastest method to compute the partition.

if zeros are allowed, then we must ask whether Q >= N. if not, then the total Q is partitioned among a greater number of parts than there are, say, individuals. In which case, some parts must be zero. A random partition would then be any random partition of Q with zeros appended at the end. But, if Q >= N, then Q is partitioned among less number of parts than there are individuals. In which case, a random partition would be any random partition of Q having N or less parts.