README.md

In leandromineti/DTScanF: Implementation a time series forecasting algorithm.

DTScanF

Overview

An R package implementing innovative forecasting algorithm for wind time series.

Installation

library(devtools)

devtools::install_github("leandromineti/DTScanF")

If you are a Windows user, make sure you have Rtools installed.

library(DTScanF)

data("curvelo", package = "DTScanF")

res <- dtsf(ts = curvelo$vento, poli = 1, best = 5, window = 24, forecast = 10)

ts_f <- cbind(matrix(NA, ncol = 24, nrow = 5), res$forecasts)
wind <- c(tail(curvelo$vento, 24), rep(NA, 10))
res_forecasting <- c(rep(NA, 24), res$forecast)  # Aggregated by median

plot(wind, col = "blue", type = "l", ylim = c(0, 3), ylab = "Wind", xlab = "Hour")
par(new=T)
matplot(t(ts_f), type = "l", ylim = c(0, 3), ylab = "", xlab = "")
par(new=T)
lines(res_forecasting, col = "orange")

leandromineti/DTScanF documentation built on May 17, 2019, 5:52 p.m.