From: thomas.e.elam@gmail.com   
      
   On 11/26/2025 11:09 AM, -hh wrote:   
   > On 11/25/25 21:12, Alan wrote:   
   >> On 2025-11-25 17:45, Tom Elam wrote:   
   >>> On 11/25/2025 3:09 PM, -hh wrote:   
   >>>> On 11/25/25 14:52, Tom Elam wrote:   
   >>>>> On 11/17/2025 1:44 PM, -hh wrote:   
   >>>>>> Yeah, so?  Recall that I eyeballed it & noted the nonlinearity   
   >>>>>> from the very start without even needing to do any math...because   
   >>>>>> I understood the basic data apparently much better than you did   
   >>>>>> (and still do).   
   >>>>>   
   >>>>> You have not responded for several days. I'm thinking that calling   
   >>>>> a simple scatter plot a statistical "work product" has to be   
   >>>>> embarrassing.   
   >>>>   
   >>>> Nope.   
   >>>>   
   >>>> You've claimed to have offered just one hour's worth of consulting   
   >>>> time in total.  What you would get for that is a summary overview of   
   >>>> what you've missed, as validated by your 'full' spreadsheet.   
   >>>>   
   >>>> Of course you'll not like that answer because you want all of the   
   >>>> work to be also done for that one hour, which is unrealistic &   
   >>>> disingenuous: the sign of a bad customer who's determined to never   
   >>>> be satisfied. Thus, you're not worth my time.   
   >>>>   
   >>>>   
   >>>>  > [snipped, without reading]   
   >>>>   
   >>>>   
   >>>> -hh   
   >>>   
   >>> OK, I'll take you up on the one hour. $150 for finding out what I   
   >>> missed. I'm always willing to learn.   
   >>>   
   >>> Of course, the advice must be useful as measured by a material model   
   >>> improvement, not duplicate what I have already done, and not call for   
   >>> additional data that is not available or does not even exist. If your   
   >>> advice calls for a statistical model beyond Excel's capability I can   
   >>> rent something else.   
   >>>   
   >>> All I need from you if the work is useful is sufficient info for a   
   >>> PayPal, Venmo or Zelle transfer.   
   >>   
   >> So you'll get to decide afterward if the "work is useful"...   
   >>   
   >> ...and then you'll pay?   
   >>   
   >> LOL!   
   >   
   >   
   > And already anticipated:   
   >   
   > "Of course you'll not like that answer... the sign of a bad customer   
   > who's determined to never be satisfied. Thus, you're not worth my time."   
   >   
   >   
   > Tommy has now confirmed that he 100% intends to welsh.   
   >   
   >   
   >   
   > -hh   
      
   I will pay if you give me useful information. You have had weeks to   
   think about this. You have stated several times that you can help   
   improve the model. It should take no more than 10 minutes to type in   
   your ideas and hit the send arrow. That's actually $900 an hour.   
      
   I do not think you have anything useful. I think you want $150 up front   
   for nothing. After all, you are the same person who told me I would need   
   to travel to England to see a friend of yours who could validate my   
   pilot logbook. You are the same person who called a simple data plot a   
   statistical model. No ethics and no expertise. But I am willing to give   
   you a chance anyway.   
      
   So what are the possibilities? Different model specification? It's going   
   to be difficult to improve on 96+% R^2. I have tried several data   
   transformations, including ln temp against ln kWh. Results were   
   marginally worse in every case. The only thing that helped was using ln   
   kWh that took out heteroscedasticity and dramatically reduced the   
   standard error of several independent variable coefficients. That was   
   all. Heteroscedasticity does not usually bias coefficient estimates, but   
   may increase their standard errors.   
      
   Different independent variables? I have tried wind and solar data, no   
   effect. Others? Help would be appreciated.   
      
   Keep in mind that just for fun I wanted to track changes in the home and   
   see if they could help account for the long term reductions seen in my   
   power use. Along with the average temperature record (below) the model   
   tracks that reduction. You saw the graph. Mission accomplished. Any   
   marginal improvement is not worth much to me. If you have something much   
   better then we can talk.   
      
   BTW, I did find one model that has a much higher R^2. The new thermostat   
   has very detailed downloadable data from which I extracted monthly run   
   time for the heat pump compressor and aux heat. I regressed that against   
   the monthly utility bill kWh hours consumed. The R^2 was over 98%. The 2   
   independent variables were all that was needed. However this is not a   
   viable regression model. Expert that you are (?) you will know the reason.   
      
   Year	Avg Temp   
   2010	54.7   
   2011	55.0   
   2012	57.0   
   2013	52.3   
   2014	49.9   
   2015	53.5   
   2016	55.4   
   2017	55.6   
   2018	53.3   
   2019	54.9   
   2020	54.9   
   2021	54.7   
   2022	53.8   
   2023	56.3   
   2024	56.9   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)