FPZIP

Abstract

fpzip is a C/C++ library for lossless and lossy compression of 1D, 2D, and 3D single- and double-precision floating-point arrays. Although designed for lossless compression, the user may optionally request lossy compression by specifying how many bits of precision per floating-point value to retain. Fpzip supports streaming compression and decompression with either a memory buffer or a file as source or destination. In addition to a simple C interface, fpzip comes with an executable for (de)compressing binary files.