Hi Eric,
as Elena's mentor, I can try to shed some light on this question.
IRLS, as used in GLM fitting, does require a matrix, but it's always
diagonal. See, for example, https://bwlewis.github.io/GLM/ , section
"Algorithm IRLS". The motivation for this pull request is exactly this
-- implement a
, 2015 at 12:29 PM, Alexey Dievsky wrote:
> Hi Gilles,
>
> thanks for your response. I've run a small JMH benchmark of both old
> (direct summation) and new (default 16KB tabulation) approaches, and here
> are the results:
>
> Benchmark
n drop the assisted
lookup altogether, reducing the memory footprint.
Sincerely,
Aleksei Dievskii.
On Thu, Nov 19, 2015 at 11:18 AM, Gilles
wrote:
> Hello.
>
> On Thu, 19 Nov 2015 10:34:46 +0100, Alexey Dievsky wrote:
>
>> Hi,
>>
>> I recently submitted an issue to JIRA (
Hi,
I recently submitted an issue to JIRA (
https://issues.apache.org/jira/browse/MATH-1293) and received an advice to
discuss it here.
In short, the current implementation of CombinatoricsUtils#factorialLog
employs direct summation and thus has linear complexity (O(n) additions and
O(n) logs). I
