etseidl commented on PR #514: URL: https://github.com/apache/parquet-format/pull/514#issuecomment-4228127480
Following up, I've modernized the #221 PoC now as well. I need to do some more testing, but early results indicate somewhere around a 5-10% speed up for floating point columns for that version, but a 3-7% slowdown for the formulation here. I have to say I still prefer the version without NaN counts. It's conceptually simpler, requires no special handling, and is indeed a good deal easier to implement. But, if we go with the NaN counts, I'm sure I can get the performance up eventually. @wgtmac do you have any thoughts either way given your experience with Java? Are we ready for a vote on this? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
