This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author mark.dickinson
Recipients mark.dickinson, martin.panter, ned.deily, rhettinger, steven.daprano, tim.peters, vstinner
Date 2016-08-17.12:31:17
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1471437078.16.0.307745785519.issue27761@psf.upfronthosting.co.za>
In-reply-to
Content
Victor: by "way too slow", I really *do* mean way too slow. :-)

floor_nroot does arithmetic with integers of bit-length approximately 54*n. For small n, that's fine, but if someone tried to take the geometric mean of a list of 50000 values (which it seems to me should be a perfectly reasonable use-case), floor_nroot would then be trying to do computations with multi-million-bit integers. The `a**(n-1)` operation in particular would be a killer for large `n`.
History
Date User Action Args
2016-08-17 12:31:18mark.dickinsonsetrecipients: + mark.dickinson, tim.peters, rhettinger, vstinner, ned.deily, steven.daprano, martin.panter
2016-08-17 12:31:18mark.dickinsonsetmessageid: <1471437078.16.0.307745785519.issue27761@psf.upfronthosting.co.za>
2016-08-17 12:31:18mark.dickinsonlinkissue27761 messages
2016-08-17 12:31:17mark.dickinsoncreate