[Radiance-general] gendaymtx

[Greg Ward]

Hi Galen,

Thanks for trying this out.  I assume the '|' symbol in your e-mail is just a stray character, otherwise you would get an error trying to execute this.

I hadn't tested this mode before with a simple daylight coefficient matrix, but it seems to work OK.  Are you sure your "daylightcoeffs.dc" file has an appropriate header to it?  The dctimestep command expects the file to start something like:

#?RADIANCE
Test matrix
FORMAT=ascii

0.954473516	0.443845774	0.2058499
0.64873052	0.152509289	0.862083965
...

Binary files will have "FORMAT=float" or suchlike, instead.  On the other hand, the sky vector or matrix file is not expected to have a header.  A bit peculiar, I know, but this is how things worked out.  Your gendaymtx has the appropriate number of lines, 876*146 plus 146 blank lines separating each sky patch row.

Cheers,
-Greg

> From: Galen Burrell <Galen.Burrell at arup.com>
> Date: March 4, 2013 4:11:50 PM PST
> 
> Pretty excited about the new 'gendaymtx' binary over here!
> 
> Hit a bit of a stumbling block on my first trial run, however.
> When I feed it an 8760 hour weather tape, it produces a file that is 3 columns wide (R G B radiance) x 1279105 rows (which seems to correlate to 8760 x 146 sky patches...almost?).  Instead of adding 3 columns for each time step, it appears to be appending each time step to the end of the file.  This creates a problem when I try to run dctimestep, which gives me a "system- header read error".
> 
> My commands are something like this:
> 
> epw2wea tacoma.epw  tacoma.wea
> rtcontrib -h- -faa -V+ -b tbin -bn 146 -m skyglow -f tregenza.cal < test.pts ${options} test.oct > daylightcoeffs.dc
> gendaymtx -m 1 -r -90 weather/tacoma.wea > weather/tacoma.smx
> | dctimestep -n 8760 daylightcoeffs.dc weather/tacoma.smx > test_illuminance.dat
> 
> Thanks for any insight on this nifty new tool.
>