Strange error while crunching

[Milk]

Hi,

i just checked one of my Genome-Clients and saw a strange error message.

10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0%jwe0217i-e
N(x) or ACOS(x), ABS(x).gt.1.0 (x=-0.100000012e+01).
error occurs at ???????? loc 0044d8dd
???????? at loc 00000000 called from loc 0044caab in ????????
???????? at loc 00000000 called from loc 0043b3e6 in ????????
???????? at loc 00000000 called from loc 0045b09a in ????????
???????? at loc 00000000 called from loc 0041ed4d in ????????
???????? at loc 00000000 called from loc 00419c02 in ????????
???????? at loc 00000000 called from loc 00414c05 in ????????
???????? at loc 00000000 called from loc 0041d7a1 in ????????
???????? at loc 00000000 called from o.s.

[????????] : Program name cannot be printed for not finding COFF symbols.

taken to (standard) corrective action, execution continuing.
100.0%

As you can see, this happened while crunching and it continued normally.
Any idea what this means?

Thx
Milk

[Dustin]

Yep, I saw that once only. The same exact thing, except for the numbers. It continued crunching, and finished the gene fine.

I have no idea what it means though.:confused:

[Medic193]

I've seen it happen more than once. In fact it continued to do it after every sequence of the gene. It's kind of strange. I think I eventually just cleared the info and started fresh.

[Bruce]

Originally posted by Milk
Hi,

i just checked one of my Genome-Clients and saw a strange error message.

10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0%jwe0217i-e
N(x) or ACOS(x), ABS(x).gt.1.0 (x=-0.100000012e+01).
error occurs at ???????? loc 0044d8dd
(snip)
taken to (standard) corrective action, execution continuing.
100.0%

As you can see, this happened while crunching and it continued normally.
Any idea what this means?

Yes, I know exactly what it means. Ignore it.

It has been discussed at length on the yahoo board and I recognize it from my years of writing fortran for geometric calculations.

It is a bug, but unimportant, so it probably won't be fixed quickly.

The standard error recovery is exactly what should happen to the calculation. And the analysis continues correctly.

When one of the angles being calculated reaches EXACTLY 90 deg., there is a roundoff error. The value (plus or minus) 1.00000012 should be rounded off to 1.0 since the true value can never exceed 1.0. The normal programming fix is either (a) inhibit the error message, or (b) test for a value that exceeds 1.0 by a neglible amount and round it off before calculating the ASIN or ACOS.

Don't clear those genes. Let them run.