Yup. This is intriguing. CC me when you hit this again, in case I miss it.

Dawid

On Wed, Nov 20, 2013 at 1:09 PM, Erick Erickson <[email protected]> wrote:
> Thanks Dawid! Problem is I can't reproduce locally But I incorporated your
> suggestion and committed it to trunk so the next time it fails we should
> have the information to go on.
>
>
> On Wed, Nov 20, 2013 at 6:30 AM, Dawid Weiss <[email protected]>
> wrote:
>>
>> > assertTrue(Math.abs(floatResult-floatTest)<.00000000001);
>>
>> I would change this to:
>>
>> assertTrue(Math.abs(floatResult-floatTest)<.00000000001 :
>>   "Oops: (double raws) " + Double.doubleToRawLongBits(floatResult) + "
>> - " + Double.doubleToRawLongBits(floatTest) + " < min.")
>>
>> this should give you raw double representation from which you could
>> then diagnose the issue (reproduce locally for the same seed, compare
>> raw repr.).
>>
>> Dawid
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to