Development of Open Forecast Support System, Stage I

The aim of the project is to develop an R Shiny application that would allow analysing data (downloaded from SQL) and produce point and interval forecasts for specified temporal and cross-sectional levels of data for the specified forecast horizon.

The Open Forecast Support System (OFSS) is focused on working with the admittance of patients in hospitals, but can be extended to work with other data. The main advantage in focusin on this data is because it has many frequencies (the data is recorded, when patients come) and many categories (sex and age groups, types of disease etc). This allows developing a system that can be easily extended and applied to other forecasting related problems, such as demand forecasting for retailers, energy forecasting, tourism forecasting etc.

The final aim of the project is to develop a OFSS constructor, which could be deployed in any organisation with a minimum knowledge of R from the admin team.

The project will be carried out in three stages:

1. The first one is to take place in 2020, and will be focused on the development of R Shiny application based on the existing forecasting packages and the connection between the R Shiny and the SQL database;
2. The second stage will focus on the development of dashboard and on the improvement of the forecasting accuracy via appropriate model selection and combination mechanisms;
3. The third stage will introduce the more complicated interactions between the users and the OFSS, allowing amending the produced forecasts and prediction intervals and add explanatory variables.

The project is done with the support from South Tees Hospitals NHS Foundation Trust in England, which will provide the data and access to the database for the purposes of development and installation of the OFSS.

The final system will be modular, allowing adding or removing modules, based on the needs of users.

---

Old stuff

This repository is a boilerplate repository that helps you prepare your proposal for the R Consortium.

Background

Set up in 2015, the R Consortium is an organisation set up to help support the R Foundation, the R Community, and R users.

The primary purpose of the R Consortium (collectively, the “Purpose”) is to:

(a) advance the worldwide promotion of and support for the R open source language and environment as the preferred language for statistical computing and graphics (the “Environment”);

(b) establish, maintain, seek support for, and develop infrastructure projects and technical and infrastructure collaboration initiatives related to the Environment, and such other initiatives as may be appropriate to support, enable and promote the Environment;

(c) encourage and increase user adoption, involvement with, and contribution to, the Environment;

(d) facilitate communication and collaboration among users and developers of the Environment, the R Consortium and the R Foundation for Statistical Computing (the “R Foundation”);

(e) support and maintain policies set by the Board; and

(f) undertake such other activities as may from time to time be appropriate to further the purposes and achieve the goals set forth above.

In furtherance of these efforts, the R Consortium shall seek to solicit the participation of all interested parties on a fair, equitable and open basis. R Consortium Bylaws, Section 1.4

Delivery of the technical aspects for R Consortium's projects is overseen by the Infrastructure Steering Committee (ISC). The ISC is set up to receive, select, and manage projects that deliver upon the aims of the Consortium. The ISC will have an ongoing call for proposals and will select proposals to move into project stage approximately every six months. Within the process notes, it does say that if a proposal is unlikely to get funded then the proposers will be notified as soon as possible, partially so that re-submission can happen in the event fixable issues.

Proposals

Here we detail useful guidance notes on making proposals to the ISC but you should always consult the ISC Proposal page as there could be updates.

• Try to complete as many of the sections of this boilerplate document as possible. Each section is included either for practical purposes or has been specifically requested by the ISC
• Add relevant additional sections, like the letter of support from an R Core member if you want a change to R itself
• Proposals should be 2-5 pages when in PDF form
• You can submit a proposal on your own, but it's really recommended to get engagement from the community (and the ISC) first
• Proposals should be emailed to [email protected]

Making your proposal

This is a boilerplate repository that you will need to fork, title appropriately and start filling in.

• Use the github fork command
• Go to the repository settings and change the name to reflect your proposal
• Create a new Rstudio project from version control and use the git URL for the repo
• Write an overview of the proposal instead of this boilerplate for the README
• Start completing the relevant Rmd pages of the proposal
• Use ghgenerate.R to build the document
• Regularly commit and push the changes to github
• Solicit feedback and contributions from others