BUG: Lux value incorrect (?)

[raptile]

Currently it is quite cloudy in holland and the lux value is "11300" .
A friend of mine has a meter and that one indicates that the value is "1266 Lux"
I think there is an extra "0" added to the actual calculated Lux value.
Could that be possible?

[oredin]

Hi,

I've checked the script. So far, it seems to me that the translation from okta to cloud coverage in percent seems good in the script.

I've looked further into the notion of okta (sky nebulosity) and cloud coverage from the intial article on domotique-info.fr. From what I've read, the two notions aren't the same things and can differ. That's why the initial contributor (Sebastien Joly) used the okta for his algorithm. The problem seems to be getting a real sky nebulosity value from a web API.

After checking the value from the nearest station on Ogimet (which was the source used to get the okta value before my changes), there is a difference.
DarkSky tells me i've got a 67% for cloud cover when Ogimet give me an okta of 9 (Sky obscured by fog and/or other meteorological phenomena) !

So, the problem isn't the calculation but the value used to do it.

I'll be looking for another source for a better okta value.