tft

This is a C# application for time series forecasting using tft's R implementation.
Through the C# UI, you can easily perform time series forecasting with the
Temporal Fusion Transformer model

At the core of this application, tft is https://github.com/mlverse.
It relies heavily on https://github.com/mlverse/tft, a wonderful library implemented in R developed at

---

Feature importance Plot

The permutation feature importance algorithm based on Fisher, Rudin, and Dominici (2018)

---

Requirements

webview2
R-4.1.2

slider
ggplot2
tidymodels
scales
dplyr
tidyverse
recipes
prodlim
listenv
rlang
plotly
htmlwidgets

setup

tft
torch
luz

.libPaths(c('Your installation path/tft/lib',.libPaths()))

GPU(cuda11.3)

Sys.setenv("CUDA_PATH"  = "C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v11.3")

CPU

Sys.setenv("CUDA_PATH"  = "")

#remotes::install_github("mlverse/torch")
install.packages("torch")

install.packages("luz", repos = "http://cran.us.r-project.org")
install.packages("slider", repos = "http://cran.us.r-project.org")
install.packages("tidymodels", repos = "http://cran.us.r-project.org")
install.packages("scales", repos = "http://cran.us.r-project.org")
install.packages("dplyr", repos = "http://cran.us.r-project.org")
install.packages("tidyverse", repos = "http://cran.us.r-project.org")
install.packages("recipes", repos = "http://cran.us.r-project.org")
install.packages("prodlim" repos = "http://cran.us.r-project.org"))
install.packages("listenv" repos = "http://cran.us.r-project.org"))
install.packages("rlang", repos = "http://cran.us.r-project.org")
install.packages("gplots", repos = "http://cran.us.r-project.org")
devtools::install_github("fisproject/lineNotify", force=T)

remotes::install_github("mlverse/tft")

tft
torch
luz