Hello,
I was looking into the benchmarking data. For the Walmart and IMDB datasets, the number of rows doesn't not match the figures you report in the paper.
I guess you refer to the subsampled version for the Walmart dataset. I get 12,075 while you report 15,317. For IMDB, the gap is pretty large, I have roughly 335K rows, while you report over 1.2M rows.
Also, could you please explain why you found the need to subsample some DBs ?
Best regards,
Antoine
Hello,
I was looking into the benchmarking data. For the Walmart and IMDB datasets, the number of rows doesn't not match the figures you report in the paper.
I guess you refer to the subsampled version for the Walmart dataset. I get 12,075 while you report 15,317. For IMDB, the gap is pretty large, I have roughly 335K rows, while you report over 1.2M rows.
Also, could you please explain why you found the need to subsample some DBs ?
Best regards,
Antoine