[ 
https://issues.apache.org/jira/browse/CALCITE-4522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17301261#comment-17301261
 ] 

Julian Hyde commented on CALCITE-4522:
--------------------------------------

[~871], Thanks for hanging in there. It must be unpleasant to have reviewers 
warring with each other, especially on your first contribution to a project.

Here is my proposed cost model:
* If fetch is zero, cpu cost is zero; otherwise,
* if there are no sort keys, cost is min(fetch + offset, inputRowCount) * 
bytesPerRow; otherwise
* cost is inputRowCount * log(min(fetch + offset, inputRowCount)) * bytesPerRow.

I think a method {{Util.nLogM(n, m)}} would be useful, where {{n}} is the 
number of input rows, and {{m}} is the number of active rows (and therefore 
determines the number of times each row is compared and/or moved). We would 
call it as follows: {{Util.nLogM(inputRowCount, fetch + offset))}}. It would 
make sure that {{m}} is at least e (and therefore log is at least 1), and make 
sure that {{m}} is no greater than {{n}}.

> Sort cost should account for the number of columns in collation
> ---------------------------------------------------------------
>
>                 Key: CALCITE-4522
>                 URL: https://issues.apache.org/jira/browse/CALCITE-4522
>             Project: Calcite
>          Issue Type: Improvement
>          Components: core
>            Reporter: hqx
>            Priority: Minor
>              Labels: pull-request-available
>          Time Spent: 9h 20m
>  Remaining Estimate: 0h
>
> The old method to compute the cost of sort has some problem.
>  # When the RelCollation is empty, there is no need to sort, but it still 
> compute the cpu cost of sort.
>  # use n * log\(n) * row_byte to estimate the cpu cost may be inaccurate, 
> where n means the output row count of the sort operator, and row_byte means 
> the average bytes of one row .
> Instead, I give follow suggestion.
>  # the cpu cost is zero if the RelCollation is empty.
>  # let heap_size be min(offset + fetch, input_count), and use input_count * 
> max(1, log(heap_size))* row_byte to compute the cpu cost.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to