Skip to contents

Abstract

Introduction

This vignette explains how to create a new workflow in the MagellanNTK framework. MagellanNTK provides a Shiny app (called Tools_Templates()) to facilitate this operation.. As there are several steps to get the information needed, the app uses MagellanNTK itself. One of the advantage is to introduce the UI of MagellanNTK

The templates that can be created are some R code with the basic (and needed) code of a framework compliant with MagellanNTK. This means that the code generated may be completed but not deleted

Features of Tools_Template()

The Shiny app is able to create three types of templates:

  • pipeline template: xxx

  • process template: xxx

  • extra module template: This template creates a module with the correct input and output to be inserted in MagellanNTK

The two first templates also create:

  • the Description markdown template

  • the generic functions that can be customized within the app

Create a workflow

MagellantNTK has a Shiny application to help developers to create a new workflow by crating a template source code. The developer has only to complete this template to achieve the code.

To launch the application, just run:

Tools_template()

This will open the Shiny app. For a complete description of the UI, please refer to xxx.

Step ‘Description’

A process is a sequence of single tasks.

A process is a sequence of single tasks.

Step ‘Create template’

A process is a sequence of single tasks.

A process is a sequence of single tasks.

Step ‘Custom dataset functions’

A process is a sequence of single tasks.

A process is a sequence of single tasks.

Step ‘Configure steps’

A process is a sequence of single tasks.

A process is a sequence of single tasks.

Step ‘Save’

A process is a sequence of single tasks.

A process is a sequence of single tasks.

Add widgets

Instantiate generic functions

Two functions are needed to modifu the list of datasets. These cannot be automatically generated by MagellanNTK because some customization may be needed and are dataset format dependant. The two functions are addDatasets() and keepDatasets

Insert and connect a module