jmeter-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aleksei Balan <abalanonl...@gmail.com>
Subject Re: Percentile and Median improvements
Date Thu, 05 Oct 2017 18:55:16 GMT
Hello Philippe Mouawad, hello Felix Schumacher,

Excuse me for returning to this already discussed topic.
I did not know about this mailing list to participate in discussion earlier.
About definition of percentile I absolutely agree with you.
Since median defined as a value dividing the tests by 2 then for even number
of tests the median can be any value between two middle values.
But in german Wikipedia it is said:
"Wenn die Anzahl der Werte gerade ist, wird der Median meist als
arithmetisches Mittel der beiden mittleren Zahlen definiert"
and in english it is also "the median is then usually defined to be
the mean of the two middle values"

In my real life experience with JMeter it is usually confusing to see
median 10 ms
for 3 successful tests of 10 ms and 3 unsuccessful of 2000 ms.

This is why my humble request to you is to agree with popular
definition of median
described in both Wikipedias.

Thank you.

Aleksei Balan


On 9/28/17, Philippe Mouawad <philippe.mouawad@gmail.com> wrote:
> Hello Aleksei,
> We reviewed your PRs. Thank you.
>
> But following analysis by different team members, we end up thinking this:
> http://mail-archives.apache.org/mod_mbox/jmeter-dev/201705.mbox/%3C96b69e93f3e23098636fa16e3be50e7f%40internetallee.de%3E
>
> What's your thoughts on that ?
>
> Thanks
> Regards
>
> On Wed, Sep 27, 2017 at 12:50 AM, Aleksei Balan <abalanonline@gmail.com>
> wrote:
>
>> Hello Apache JMeter community,
>>
>> Can you please review these two pull requests
>> https://github.com/apache/jmeter/pull/295
>> https://github.com/apache/jmeter/pull/296
>>
>> They are fixing some calculation defects. I understand that these errors
>> affect the results very rarely (especially on big number of samples)
>> but still I think it is a good idea to have more precise calculations.
>>
>> Thank you.
>>
>> Aleksei Balan
>>
>

Mime
View raw message