Time Series Forecasting Using Nearest Neighbors

Allows to forecast time series using nearest neighbors regression Francisco Martinez, Maria P. Frias, Maria D. Perez-Godoy and Antonio J. Rivera (2017) . When the forecasting horizon is higher than 1, two multi-step ahead forecasting strategies can be used. The model built is autoregressive, that is, it is only based on the observations of the time series. The nearest neighbors used in a prediction can be consulted and plotted.

The goal of tsfknn is to forecast time series using KNN regression


You can install tsfknn from from github with:

# install.packages("devtools")


This is a basic example which shows you how to forecast with tsfknn:

pred <- knn_forecasting(USAccDeaths, h = 12, k = 3)
pred$prediction # To see a time series with the forecasts
plot(pred) # To see a plot with the forecast
autoplot(pred, highlight = "neighbors")  # To see the nearest neighbors


tsfknn 0.2.0

  • summary and print.summary methods are added for "knnForecast" objects
  • String parameters are processed with math.arg
  • Fix calculation of how many KNN examples has the model in knn_forecasting
  • Weighted combination of the targets of nearest neighbors is implemented
  • A function that computes the number of training instances that would have a model has been added

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


0.5.0 by Francisco Martinez, 7 months ago


Report a bug at https://github.com/franciscomartinezdelrio/tsfknn/issues

Browse source code at https://github.com/cran/tsfknn

Authors: Francisco Martinez [aut, cre]

Documentation:   PDF Manual  

Task views: Time Series Analysis

GPL-2 license

Imports ggplot2, graphics, Rcpp, stats, utils

Suggests knitr, rmarkdown, testthat

Linking to Rcpp

See at CRAN