Same results from radiance cumulative sky study

Hi Ladybug community!
I’ve lately run into this problem. I’m trying to run three cumulative sky studies for winter/summer and the whole year using Radiance. The problem is that I get the same values when I run these studies simultaneously. Simply changing working directories and file names does not resolve the issue.
I’m suspecting that the problem lies in the cumulative sky generation as all three generated *.sky files reference the T_1.cal file that contains the sky cumulative radiation values. Shouldn’t there be three different *.cal files, one for each generated sky instead of one?
I could simply run one sim at a time, however in my case I’m doing a parametric study and this is impossible. Is this a bug? Can you suggest a workaround?
I’m attaching the gh file that illustrates the problem.

Aris

same_rad_results.gh (507.5 KB)

To partially answer my own question:
I rerun GenCumSky for each time period and manually renamed the *.cal files. Then I edited the *.sky files to reference the correct *.cal files for each period. This seems to work for now and I get reasonable results but the problem needs to be addressed in a future release.

Hi @mostapha,

Just want to bump this issue up again as I ran into it today.
Would you think it make sense to have different *.cal files for each analysis period?

genCumSky component

# calFile = subWorkingDir + "\\" + newLocName + '_1.cal' # original

#write the sky file
# read the analysis period to name the file
stMonth, stDay, stHour, endMonth, endDay, endHour = lb_preparation.readRunPeriod(analysisPeriod, False)

calFile = subWorkingDir + "\\" + newLocName + '_1_%s.cal' \
% "_".join([str(stMonth), str(stDay), str(stHour), str(endMonth), str(endDay), str(endHour)])

genCumSkyStr() in Ladybug_Ladybug

        # calFileName = subWorkingDir + "\\" + newLocName + '_' + sunMode[-1] + '.cal' orignal
        calFileName = subWorkingDir + "\\" + newLocName + '_' + sunMode[-1] + '_%s.cal' \
        % "_".join([str(stMonth), str(stDay), str(stHour), str(endMonth), str(endDay), str(endHour)])

It has to be different .cal files for each period. The cumulative data used to assign radiation values to the sky patches are derived from the numbers that are hard coded into the .cal file. Those numbers will change based on the time intervals assigned.

Hi @vhoang and @ArisVartholomaios, I have to take a closer look to remember which code is used for what. Generally speaking it shouldn’t be a problem in Ladybug as the generated sky is always annual and the select sky component does post-process the annual values for a certain analysis period.

In Honeybee this can be an issue but if I understand correctly it does already rename the cal file based on the analysis period?

Hi @mostapha,

As far as I can see, in current Honeybee the cal file is not named based on analysis period but rather being overwritten for each period. The cumulativeSky.bat always outputs *_1.cal for all periods of the same epw file.

The snippets in my previous post were my workaround by renaming the file based on period instead. Not sure if this would cause problems with overall HB workflow but for my grid based solar radiation analysis it helps avoiding the issue of wrong sky condition.

Hi @vhoang, Then that’s a bug! I can update the code but I wonder if you can send a pull request to GitHub? That will be a :1st_place_medal: