DelayedRandomArray

DelayedRandomArray implements a DelayedArray of random values where the realization of the sampled values is delayed until they are needed, enabling reproducible sampling within any subarray through chunking with different random seeds and streams.

Key Features:

• Delayed realization: Random values are not generated until they are explicitly requested, conserving memory for large arrays.
• Reproducible subarray sampling: Each chunk is initialized with a different random seed and stream to ensure reproducibility within any subarray.
• Distribution support: Supports the usual distributions in the stats package, with scalar, vector, and array parameters.

Scientific Applications:

• Large-scale statistical simulation: Enables generation of massive random arrays without immediate memory allocation for Monte Carlo analyses.
• Reproducible computational biology: Supports reproducible random sampling in bioinformatics pipelines that operate on large genomic or biological datasets.

Methodology:

Chunking divides the array into chunks, each initialized with a different random seed and stream to achieve reproducible sampling within any subarray.