Bit Grooming: statistically accurate precision-preserving quantization with compression, evaluated in the netCDF Operators (NCO, v4.4.8+)

Zender, Charles S.

doi:https://doi.org/10.5194/gmd-9-3199-2016

Articles | Volume 9, issue 9

https://doi.org/10.5194/gmd-9-3199-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/gmd-9-3199-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 9, issue 9

Development and technical paper

|

19 Sep 2016

Development and technical paper |

| 19 Sep 2016

Bit Grooming: statistically accurate precision-preserving quantization with compression, evaluated in the netCDF Operators (NCO, v4.4.8+)

Charles S. Zender

Download

Final revised paper (published on 19 Sep 2016)
Supplement to the final revised paper
Preprint (discussion started on 21 Apr 2016)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Reviewer comments', Anonymous Referee #1, 19 May 2016
- AC1: 'Response to Reviewer #1', Charles Zender, 31 May 2016
RC2: 'Review Comment', Anonymous Referee #2, 11 Jul 2016
- AC2: 'Response to Reviewer #2', Charles Zender, 30 Jul 2016

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Charles Zender on behalf of the Authors (22 Aug 2016) Author's response Manuscript

ED: Publish subject to minor revisions (Editor review) (30 Aug 2016) by David Ham

AR by Charles Zender on behalf of the Authors (01 Sep 2016) Author's response Manuscript

ED: Publish as is (02 Sep 2016) by David Ham

AR by Charles Zender on behalf of the Authors (02 Sep 2016) Manuscript

Short summary

We introduce Bit Grooming, a lossy compression algorithm that removes the bloat due to false precision, those bits and bytes beyond the meaningful precision of the data. Bit Grooming is statistically unbiased, applies to all floating-point numbers, and is easy to use. Bit Grooming reduces data storage requirements by 25–80 %. Unlike its best-known competitor Linear Packing, Bit Grooming imposes no software overhead on users, and guarantees its precision throughout the whole floating-point range.