alamb commented on PR #195: URL: https://github.com/apache/parquet-site/pull/195#issuecomment-5217715155
I think one of the important things we should do for this blog is to produce some performance charts to compare ALP Specifically I think we should 1. Pick representative data sets (also we should find one that has actual floating point data that ALP doesn't work for) 2. Encoding/compression: PLAIN encoding + no compression, PLAIN Encoding + zstd, ALP 4. For each data set, for each compression comparison, report (compression speed (GB/s), decompression speed (GB/s), and compressed size (bits/value)) We can follow the model of the original paper, but compare against existing parquet encodings, not other techniques https://ir.cwi.nl/pub/33334/33334.pdf <img width="384" height="512" alt="Screenshot 2026-08-07 at 9 30 45 AM" src="https://github.com/user-attachments/assets/ecaded7a-edd5-4bc1-bc5a-7053bb7efd20"; /> <img width="364" height="247" alt="Screenshot 2026-08-07 at 9 30 54 AM" src="https://github.com/user-attachments/assets/24b6e709-187f-4654-a84b-434eed54a2cf"; /> -- 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]
