ALP: Adaptive lossless floating-point compression (github.com)

<a href="https://news.ycombinator.com/item?id=49051355">Comments</a>