Palmer Penguins Dashboard Explorer

PPDE (Palmer Penguins Dashboard Explorer) is an interactive R Shiny dashboard for exploring the Palmer Penguins dataset through a CRISP-DM-inspired workflow.

The repository combines a compact teaching example with a more complete exploratory dashboard. It can be used during classes to demonstrate how a Shiny application grows from a minimal prototype into a richer analytical tool.

Repository contents

.
├── README.Rmd
├── pp-quick-start.Rmd
└── ppde.R

pp-quick-start.Rmd

A step-by-step introduction to R Shiny designed for classroom use.

The document demonstrates how to build a simple Shiny application incrementally, including:

• application structure,
• user interface and server logic,
• reactive inputs and outputs,
• basic visualisations,
• filtering and interaction.

ppde.R

The complete PPDE dashboard.

The application includes:

• dataset overview and description,
• summary statistics and data preview,
• histograms, density plots and violin plots,
• correlogram,
• missing-data assessment,
• potential outlier detection,
• interactive 2D and 3D scatterplots,
• colour and marker-shape mappings,
• trend lines for 2D plots,
• island map with species filtering,
• decision tree classification,
• confusion matrix,
• accuracy, balanced accuracy and per-class evaluation,
• light and dark modes,
• high-contrast mode,
• a colour-blind-safe palette.

CRISP-DM-inspired workflow

The dashboard loosely follows selected stages of the CRISP-DM process.

1. Data understanding

• dataset description,
• descriptive statistics,
• data preview,
• distribution plots,
• correlation analysis.

2. Data quality assessment

• missing-value analysis,
• potential outlier detection.

3. Data exploration

• interactive 2D and 3D visualisation,
• colour and marker-shape mappings,
• species and island filtering,
• trend analysis.

4. Geographical context

• island-level summaries,
• species distribution across islands,
• interactive map popups.

5. Modelling and evaluation

• decision tree classifier,
• test-set predictions,
• accuracy,
• balanced accuracy,
• confusion matrix,
• per-class precision, recall and F1 score.

The application is intended as an educational walkthrough rather than a complete production implementation of CRISP-DM.

Requirements

Install the required R packages:

install.packages(c(
  "shiny",
  "shinythemes",
  "palmerpenguins",
  "ggplot2",
  "dplyr",
  "tidyr",
  "plotly",
  "leaflet",
  "naniar",
  "scales",
  "rpart",
  "rpart.plot"
))

Run the dashboard

Clone or download the repository, set the repository directory as the working directory and run:

shiny::runApp("ppde.R")

Alternatively, open ppde.R in RStudio and select Run App.

Dataset

The application uses the penguins dataset from the palmerpenguins R package.

The dataset contains observations of three penguin species from the Palmer Archipelago, Antarctica:

• Adelie,
• Chinstrap,
• Gentoo.

The observations come from three islands:

• Biscoe,
• Dream,
• Torgersen.

The available variables include:

• species,
• island,
• bill length,
• bill depth,
• flipper length,
• body mass,
• sex,
• year.

Educational use

The repository can be used to:

• introduce the structure of Shiny applications,
• demonstrate reactive programming,
• explain exploratory data analysis,
• discuss visual encoding with colour and marker shape,
• compare 2D and 3D visualisations,
• introduce data-quality assessment,
• demonstrate a basic machine-learning workflow,
• discuss accessibility in interactive data applications,
• show how an application can follow a CRISP-DM-inspired analytical process.

Project status

This repository contains material prepared for teaching and demonstration purposes.

The dashboard may be extended in the future with additional data-cleaning operations, modelling methods, deployment options and reusable modules.

Live demo

A live deployment is available on shinyapps.io:

Open PPDE

If the service is temporarily unavailable, run the application locally:

shiny::runApp("app.R")

Licence

This project is shared under the:

Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
CC BY-NC-ND 4.0

The material may be shared with attribution for non-commercial purposes. Modified versions may not be distributed under this licence.

Project Anatomy

Head  (concept & design):           MagT        ,_,
Eyes  (exploration & insight):      ambr0wl    (O,O)
Hands (development):                MagTk      (   )
Feet  (reality check & grounding):  anovi      -"-"-
>      Curious?     --------->      https://magt.ovh