The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

Getting started with proteus

Giancarlo Vercellino

proteus fits a variational sequence-to-sequence model to one or more time features and returns forecasts, uncertainty summaries, diagnostic plots, and error metrics. Version 2.0 keeps the neural network and plotting dependencies small; optional packages are loaded only when their feature is requested.

A first forecast

The package includes amzn_aapl_fb, a data frame with daily prices and a date column. A compact run is:

library(proteus)
fit <- proteus(amzn_aapl_fb, target = "AMZN", dates = "Date",
               past = 30, future = 10, epochs = 5,
               future_plan = "future::sequential", verbose = FALSE)
fit$prediction$AMZN

The prediction table contains quantiles, location and scale summaries, and distribution diagnostics. fit$plot$AMZN visualizes the historical series and forecast interval, while fit$features_errors reports back-test metrics.

Optional features

Set smoother = TRUE or use omit = FALSE with missing values to opt into fANCOVA or imputeTS, respectively. Parallel cross-validation can be enabled with future_plan = "future::multisession" after installing future and furrr. The default sequential plan works with the core dependencies.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.