GenCumulativeSky - ignore diffuse irradiance


I'm trying to run GenCumulativeSky and ignore diffuse irradiance, but
when I use the -d parameter I get an "invalid input parameter '-d'"
error. Is this a bug or is it expecting something else?

Thank you,


Hi all,

I have the same issue. Does anyone have any suggestions?


You are replying to a very old post, which should have been placed in radiance-daysim when it was created, since GenCumulativeSky is part of that distribution (I believe).

I do not know who maintains this tool, if indeed it is still maintained. You might consider using gendaymtx with the -A option as an alternative.


Thanks for the quick reply.

That indeed seems like a good idea, since GenCumulativeSky seems to be no longer supported as far as I can tell. This leads me to a second question (about gendaymtx, so I hope I am in the right place this time ;).

I was previously using GenCumulativeSky to generate a cal file that could then be used to define the cumulative sky. This was used with oconv / rtrace to calculate annual irradiance at center points of a set of windows. The results calculated by rtrace were in KWh/m^2/year, which is what we needed.

So now I am wondering how to achieve the same with gendaymtx? I had a look at the gendaymtx PDF, but I am not clear on which options to use.

I have tried running “gendaymtx -m 1 -A -c 1 1 1 singapore.wea > singapore.mtx”, but I am not sure if this is correct. I understand the gendaymtx results to be in W/sr/m^2, so I am also wondering how I would be able to get the rtrace results to be KWh/m^2/year.

Hi Patrick,

Yes, the command you’re using will generate a sky vector corresponding to the average W/sr/M^2 of each patch. This is most easily applied in a daylight coefficient calculation, where you’ve used rfluxmtx to calculate the sky-to-point coefficients and dctimestep multiply the matrix against the gendaymtx sky vector output. (See this presentation if you need basic pointers, starting at slide 38.)

The alternative is to convert the output to a .dat file and use it with a colordata (or brightdata) primitive with or similar, but this is a bit less standard. In the end, you should get similar results from rfluxmtx that you would from rtrace.

In either case, you would multiply the final result by 8760/1000 to get KWh/m^2/year, assuming the -I+ option was used and your weather file contains evenly-spaced values.


Thanks, I will look into this and see if I can figure it out.