Software

Open sources packages for R

<糖心Vlog>Our R Packages

Packages maintained by Centre for Marketing Analytics and Forecasting members:

  • <糖心Vlog class="title"> greybox

    Functions and instruments for regression model building and its application for forecasting.

  • <糖心Vlog class="title"> smooth

    Functions implementing Single Source of Error state space models for purposes of time series analysis and forecasting.

  • <糖心Vlog class="title"> diffusion

    Various diffusion models to forecast new product growth.

  • <糖心Vlog class="title"> legion

    Functions implementing multivariate state space models for purposes of time series analysis and forecasting.

  • <糖心Vlog class="title"> tsutils

    Various tools useful for time series analysis and forecasting.

Features

greybox package

greybox is the R package that is developed for marketing analytics and for forecasting using regression models.

It includes a variety of analytical tools: spread(), assoc() and others). They allow measuring and visualising associations between variables of different types.

alm() function implements regression with a variety of distributions and focuses on forecasting.

stepwise() function implements a trace forward stepwise selection based on information criteria.

lmcombine() produces regression models combined from a pool of models via AIC weights.

Read more about greybox at and the .

The package is maintained by Ivan Svetunkov.

To install the package, run this command in R:

install.packages("greybox")

smooth package

The package implements Single Source of Error state space models, which are used in forecasting. This includes:

  • ETS in es() and adam()
  • ETSX in es() and adam()
  • ETS for intermittend demand via adam()
  • ARIMA via ssarima() and adam()
  • Multiple seasonal ETS via adam()
  • Multiple seasonal ARIMA via msarima(), auto.msarima() and adam()
  • Complex Exponential Smoothing via ces()
  • Classical seasonal decomposition for multiple seasonal series
  • Simple Moving Average
  • Functions to simulate data from ETS, ARIMA, CES, SMA and other models.

Read more at and the .

The package is maintained by Ivan Svetunkov.

To install the package, run this command in R:

install.packages("smooth")

legion package

The package legion implements multivariate models in Single Source of Error state space framework. It currently has the following models:

  • ves() - Vector Exponential Smoothing, supporting a variety of options for persistence, transition and measurement matrices;
  • vets() - Vector ETS, supporting pure additive and pure multiplicative models with PIC restriction;
  • auto.ves() and auto.vets() - functions for automatic selection of VES/VETS components.
  • sim.ves() - function simulations data based on VES model.

The package is maintained by Ivan Svetunkov and Kandrika Pritularga.

To install the package, run this command in R:

install.packages("legion")

diffusion package

Package implements various diffusion models to forecast new product growth. It currently supports:

  • Bass model,
  • Gompertz model,
  • Gamma/Shifted Gompertz model,
  • Norton-Bass model

The package is maintained by .

To install the package, run this command in R:

install.packages("diffusion")

tsutils package

Package implements a variety of tools useful for time series analysis and for forecasting.

  • abc(), xyz() and abcxyz() - functions helping in categorising products by their profitability and predictability,
  • decomp() - function for classical time series decomposition with additional features,
  • nemenyi() - function for statistical comparison of performance of different forecasting techniques,
  • seasplot() - seasonal plots with advanced functionality, supporting a variety of plotting capabilities and time series components tests,
  • theta() - forecasting method by , implemented in their original formulation.

The package is maintained by .

To install the package, run this command in R:

install.packages("tsutils")

Behind the scenes

People behind the Centre for Marketing Analytics and Forecasting packages:

Frequently Asked Questions

<糖心Vlog id="accordion-heading-548195" class="hide"> accordion

<糖心Vlog>Software surveys

Our Centre has conducted surveys tracking the situation in the marketing of Forecasting Software. These surveys were published in Operational Research and the Management Sciences (ORMS) Today magazine:

  • 2022:
  • 2020:
  • 2018:
Grow with us
Work with us
Services