Automatic Source-to-Source Error Compensation of Floating-Point Programs | IEEE Conference Publication | IEEE Xplore