Hi all,

Thank you so much for your effort and above all for DIVA GH.

I'm trying to understand more about DIVA (i bought two years ago Daylighting Handbook I as well, when vol.II?), but this morning i noticed an error (maybe my fault or a bug? my DIVA v 4.0.2.39 64 bit), when i set "window component" (i refer to this tutorial and files https://vimeo.com/178178194) with deafult options (same of tutorial 05) the "Annual Daylight component" can't run. This is error message:

--------------------------

1. Write failed.
4 manual shades
0 auto shades
manual sensor ray count = 2409
Pipe terminata.

Pipe terminata.

Pipe terminata.

Pipe terminata.

Pipe terminata.

Pipe terminata.

Pipe terminata.

raytrace loop encountered a break at 0

-----------------------------

I noticed when i set windows component on Shade Geometry>NoShade it runs fine (while on RollerShade_Autogenerate i've the above error. I hope someone can help me.

Thank you in advance.

Tags: 05, dynamic, issue, shading, tutorial

Views: 240

Reply to This

Replies to This Discussion

Hi Luciano,

I don't wee anything wrong with your sceen capture. Can you post the .gh file so I can take a closer look?

Thanks,

Jon

Firstly thank you for your answer. I will attach both files 3dm + gh.

Attachments:

Definition runs fine on my end.  The only things I can't replicate on my machine are:

1) The number of threads.  Can you reduce the Number of Processes to 2 and see if the same error occurs?

and

2) Culture info. Sometimes Radiance and Daysim get thrown off by non-period decimal separators, though I don't know why this would depend on the presence of shades.  Anyway, you could try changing your decimal char to a period if it isn't already ("additional settings" under Control Panel > Region).

Thank you Jon for your help.

I run simulation reducing the number of Processes to 2 but nothing, so i have tried to set the number to 1 and it seems to run fine, but it takes some minutes.

However i attach my "additional settings" as well, may be something it could be wrong. Thanks again.


Jon Sargent said:

Definition runs fine on my end.  The only things I can't replicate on my machine are:

1) The number of threads.  Can you reduce the Number of Processes to 2 and see if the same error occurs?

and

2) Culture info. Sometimes Radiance and Daysim get thrown off by non-period decimal separators, though I don't know why this would depend on the presence of shades.  Anyway, you could try changing your decimal char to a period if it isn't already ("additional settings" under Control Panel > Region).

Attachments:
Right, your decimal separator is a comma -- did you try changing it to a period?

You are right, it works.

I have to remember to switch my decimal separator from comma to period in that case. Thanks!

Jon Sargent said:

Right, your decimal separator is a comma -- did you try changing it to a period?

Okay, that makes sense.  DIVA actually forces the decimal separator to a period on the main thread, but I must have neglected to enforce this when the job splits during multi-state runs (shades up, shades down, etc.).  Thanks for the report, I'll try to clean this up in future revisions.

Hi, several of my students experienced the same problem. For most of them, the problem was solved by reducing the amount of CPUs used for the simulation. There is however one students where I both tried to lower the amount of CPUs used (and is currently at 1) and where I changed the decimal separator to period.

Any idea how to solve this? The script works fine on my own computer.

/Jouri

Indeed, it seems I hadn't (yet) kept my word on the fix.  I just updated the source code, so multi-processor runs should work out of the box as of the next public revision.  As for the one intransigent machine... I'm afraid I don't have any good guesses at the moment.  :/

Hi Jon, i've the same problem with the Annual Analisis for the Manual Shade, because with the number of Processes to 2 it's the same grid if i haven't the shades. I can do nothing to resolve this problem?

Because if i try to put the process on a number much big that 2 tha analisis no run!

P.S:I'va already change my decimal separator in a period. 

Thank you and sorry.

Jon Sargent said:

In effetti, sembra che non abbia (ancora) mantenuto la mia parola sulla correzione. Ho appena aggiornato il codice sorgente, quindi le esecuzioni multiprocessore dovrebbero funzionare immediatamente dopo la revisione pubblica successiva. Per quanto riguarda l'unica macchina intransigente ... temo di non avere buone ipotesi al momento. : /

Reply to Discussion

RSS

© 2018   Created by jeff niemasz.   Powered by

Badges  |  Report an Issue  |  Terms of Service