search

faradayio / lodging

A carbon model
GNU Affero General Public License v3.0
2 stars 2 forks source link

Strange pool count threshold #3

Closed rossmeissl closed 12 years ago

rossmeissl commented 12 years ago

In the impact model we have:

unless characteristics[:property].pools_indoor.nil?
  [characteristics[:property].pools_indoor.to_f, 5].min
end

Where's this 5 coming from? If this is some way to guard against "99"s in the data, we should be fixing this in the DM block, not here. Impact models shouldn't have to worry about the validity of CM1's reference database.

@seamusabshere you could take this one too if you'd rather do it.

ihough commented 12 years ago

Lauren is on the case investigating unusual numbers of pools.

That said I think of this as less trying to filter out bad data and more recognizing the bounds within which our model is applicable. Right now our estimate of pool energy use is pretty rough, so even if a property really does have 25 pools accounting for just 5 of them may well give a more accurate estimate.

rossmeissl commented 12 years ago

Ah, OK, fair enough if that's the reason