Radiance Paths - a solution

I have just run a search of the code on the Python code Honeybee_Honeybee.py looking for Radiance.

There seems to be a very small number of places where the location of Radiance is hard-wired into the code. Given that NREL are releasing up-to-date versions of Radiance for Windows as they are released for Unix, and these have been placed into C:\Program Files (x86)\Radiance as standard, and this is an easily locatable (via PATH settings) standard location, could this requirement to have Radiance installed in a hard-wired location be removed?

Radiance_Paths_HoneyBee.txt

Replies are closed for this discussion.

Replies to This Discussion

Permalink Reply by Chris Mackey on September 24, 2015 at 4:26pm

Michael,
This seems like a reasonable request. If I am not mistaken, I think that Honeybee currently checks the PATH settings and, if it doesn't find radiance there or if it accidentally finds Radiance installed from DIVA (which has a different structure), it checks off the C drive. At the least, I imagine that we can check the location in ProgramFiles if it does not find an installation off of the C drive. Let's hear what Mostapha has to say and, if this method is good by him, I can implement it.

-Chris

Permalink Reply by Sarith Subramaniam on September 26, 2015 at 3:24pm

Echoing much of what Chris mentioned,having Python get access to the Radiance binaries and library files by looking into PATH and RAYPATH (through os.environ I assume) is a great idea if one can ensure that python gets to the binaries distributed by NREL.

However, a lot really depends on the order in which Radiance, DIVA, DAYSIM etc. were installed. I have run into issues with Daysim in the past, so much so that I now start DAYSIM with a batch file that first sets the PATH and RAYPATH variables before launching the Daysim interface.

Permalink Reply by Michael Donn on September 28, 2015 at 3:25am

I agree that with a single user on a single computer the installation placement of Radiance can cause issues, but I think that the DAYSIM issues are more to do with the flexible data structures that Radiance allows and the different ways in which Materials can be handled. When these meet a more rigid structure then I recollect that when the calculation is implemented with one of the alternative structures Radiance can be left with no way to find the Material definitions file.

I think this is what one of the notes in the Gracehopper code references...

Permalink Reply by Mostapha Sadeghipour Roudsari on October 1, 2015 at 7:55pm

Thanks Michael,

As Chris and Sarith mentioned Honeybee searches for radiance installation folder on the system but it only will find the folder if the path is already system paths:

https://github.com/mostaphaRoudsari/Honeybee/blob/master/src/Honeyb...

The reason to avoid installation with DIVA is mainly their customized folder structure for Daysim and Radiance which can cause issue.

There is one extra check to make sure the path doesn't have white space (https://github.com/mostaphaRoudsari/Honeybee/blob/master/src/Honeyb...). This is my fault that Honeybee currently can't handle path with white spaces. It will be fixed in the near future (https://github.com/mostaphaRoudsari/Honeybee/issues/13).

In any case the first two portion is a bug and should not be hard-coded. I just fixed them and uploaded the new version to github. Thanks!

Mostapha

RSS