I'm a noob to the group and am after statistical advice. Any interest would be really great. I expect to use R for analysis and have already read the below page which demonstrates correlations for energy use against NOAA temperatures. I suspect my problem is going to be harder to analyse.

I will run an experiment later this year. After that:

I should have six months of half hourly readings of home energy use for 30 homes (6 controls). These are kWh smart meter floating point values. At specific times during the sample (such as every few days), the treatment home occupants will get an informational stimulus which may affect their electricity consumption. Perhaps this effect happens soon after the event (such as minutes) but it may be days. The effect should be a drop in energy consumption, maybe sustained, perhaps not.

The below R example is for NOAA weather data as it correlates to energy use. The writer finds strong correlations and I suppose this is a much easier correlation to hunt for. I need to say that certain quasi regular interventions had some sort of signature in the energy use time series. I suspect I already need weather series (and control series) to help in pre-processing the data.


Any advice about directions or things to try would be very gratefully received. Cheers Craig


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.