Subsampled Riemannian trust-region (RTR) algorithms

Injecting image priors into Learnable Compressive Subsampling

This C script subsamples raw FASTQ files (single/paired-end). For paired-end, it synchronously selects read pairs from two files, ensuring retention of matched sequences. It reads input, randomly chooses entries via probability, writes outputs. Handles large files efficiently, maintains FASTQ structure, uses command-line args for input/output paths

Create a model to predict if a customer will leave the bank.

Subsample reads from a fastq/fasta file to a desired read depth. Choose random selection or new mean read length