In antononcube/QRMon-R: Quantile Regression Monad

library(QRMon)
library(splines)
library(quantreg)
library(purrr)
library(magrittr)
library(ggplot2)

Temperature prediction

In this sub-section we compute the conditional distribution of tomorrow's temperature given today's temperature.

weVec <- dfTemperatureData$Temperature
qDF <- data.frame( YesterdayValue = weVec[-length(weVec)], TodayValue = weVec[-1] )
qs <- c(0.01,seq(0.1,0.9,0.1),0.99)
qrTempObj <- 
  QRMonUnit( setNames( qDF, c("Regressor", "Value") ) ) %>% 
  QRMonQuantileRegression( df = 6, degree = 3, probabilities = qs )

qrTempObj <- qrTempObj %>% QRMonPlot

Here is an example of a prediction for tomorrow's temperature given that the temperature today is [8^{\circ}C].

res <- qrTempObj %>% QRMonPredict( newdata = c(8) ) %>% QRMonTakeValue
res <- setNames( dplyr::bind_rows( res, .id = "Quantile" ), c("Quantile", "TodayTemperature", "TomorrowTemperture") )
res$Quantile <- as.numeric(res$Quantile)
res

qrTempObj2 <- qrTempObj %>% QRMonConditionalCDFPlot(8, echoQ = F)
(qrTempObj2 %>% QRMonTakeValue) +
  ggplot2::geom_vline( xintercept = 8, color = "blue" )