Empty EPW file


#21

@KatsuyaObara

Thanks for bringing this up, this was a bug that I’m glad we caught early. The error was occurring due to, as I referred to earlier, differences in how matlab and python takes the log of negative numbers. In such cases, python throws an error, whereas matlab returns a complex number. However, the code shouldn’t be calculating negative logarithms at all, and the the correction had to occur further upstream in the RSM.VDM method, where wind speed magnitudes are calculated with respect to reference building heights (at 2m) like this:

image

y = wind speed, x = building height

This is definitely a bug, as wind speeds here are approaching negative infinity the closer building heights get to a 40 meter reference height. The only part of the code where this appears is actually within a section that is listed for experimental code, and doesn’t actually appear in the actual UWG calculation. So I was able to safely comment it out.Outside of that, the RSM.VDM method is only be used for rural sites where heights are never going to get above 40m.

Anyway, I’ve corrected the code. If you sync with the latest commit I believe your 51 m building shouldn’t cause any problems.