In EcotopeResearch/rterm: Temperature-Energy Regression Models for Building Energy Use Data

Constructing a term

# These keys have been registered for the sole use of building rterm vignettes
google_key <- "AIzaSyCOavbshIM5-TY3Dm07ar_l_BwR6gvQYWk"
noaa_key <- "fBQbabTcSnILcGQWxfXBYLMclnUKLVlH"

Everything goes through a "term" S3 class. It may seem a bit complicated up front, but mastering the "term" methods makes this all much easier. We'll walk through an example for the Ecotope office bills, step-by-step, then put it all together at the end.

Initialize

Use the constructor function newTerm() to initialize a new temperature-energy-regression-model.

library(rterm)
term <- newTerm()

Add data

Add data to the term with the addData() function.

data(ecotope)
head(ecotope)
term <- addData(term, ecotope, kwhd ~ dateStart + dateEnd)
term

There are several ways to add data with the addData() function. Above is a logical one for the ecotope dataset. Here are some more. If you try to add data to a term that already has data, it will overwrite and issue a warning.

term <- addData(term, ecotope, kwhd ~ dateStart + days)
term <- addData(term, ecotope, energyVar = "kwhd", dateEndVar = "dateEnd", daysVar ="days")

Add Weather

There is an addWeather() function to associate a weather file with the energy use dataset. In this case you can actually add multiple weather files, that will all get evaluated. This can be useful if you are unsure about which weather station to use, and want to know the consequences of the choice on inference.

The two basic choices here are to either use the NOAA API from within rterm, or provide your own weather data. In this example we will do both. Note that the NOAA API can be pretty slow. For custom weather we will use the rterm package dataset "hourlyWeather", which is approximately 10 years worth of hourly NOAA QCLCD weather (the API provides GHCN data) from Sea-Tac Airport.

stationSearch("seattle")
term <- addWeather(term, stationid = "GHCND:USW00094290", name = "Sand Point")
data(hourlyWeather)
term <- addWeather(term, hourlyWeather, temp ~ time, name = "Sea-Tac Hourly")
term

Add TMY

You can also associate TMY files with a term.

tmySearch("Seattle, WA")
term <- addTmy(term, c("WASeattle3.tm2", "WASeattleBoeing3.tm2"))

Add Methodologies

Use the addMethod() function to associate methodologies to the term.

term <- addMethod(term, "change-point", heating = TRUE, cooling = FALSE)
term <- addMethod(term, "degree-day", heating = TRUE, cooling = FALSE)
term

Evalute

term <- evaluate(term)
term

Altogether now

I recommend using the magrittr/dplyr pipe to set up the term in a single, hopefully readable statement as follows.

library(dplyr)
term <- newTerm() %>%
  addData(ecotope, kwhd ~ dateStart + dateEnd) %>%
  addWeather(stationid = "GHCND:USW00094290", name = "Sand Point") %>%
  addWeather(hourlyWeather, temp ~ time, name="Sea-Tac Hourly") %>%
  addMethod("changepoint", heating = TRUE, cooling = FALSE) %>%
  addMethod("degreeday", heating = TRUE, cooling = FALSE) %>% 
  addTmy(c("WASeattle3.tm2", "WASeattleBoeing3.tm2")) %>%
  evaluate()
term

EcotopeResearch/rterm documentation built on Oct. 17, 2022, 4:02 p.m.