http://www.perlmonks.org?node_id=997211


in reply to Specialized data compression

Thanks for posting such an interesting question!

Another idea I've seen work is to transpose your CSV, so that the first row is the timestamp, the second row is the X, etc. Then compress this. As the X's may be self-similar, and the Y's self-similar, but the X's and Y's are not, this may give the compression algorithm a break.

The timestamp should also be delta, except the first reading on resume.

-QM
--
Quantum Mechanics: The dreams stuff is made of