🐶 CheemsPlot: Interactive Dog-Styled Boxplot Explorer

An interactive Shiny dashboard that transforms traditional boxplots into a playful data visualization experience by replacing outliers with customizable 🐕 dog icons, powered by the ggdogs package.

Built as a single-file Shiny app (app.R), designed to run locally, live on GitHub, and deploy to shinyapps.io — with a clear migration path to Posit Connect Cloud.

---

🚀 Features

• 📂 Upload your own CSV — no hardcoded datasets required
• 🔄 Dynamic variable selection — categorical and numeric columns are auto-detected
• 📊 Boxplots on the fly — computed reactively as you change inputs
• 🐶 Dog-styled outliers — replace IQR outliers with 15+ dog icons
• 🎛 Full customization — dog style, dog size, orientation, and plot title
• 🖼 Horizontal or vertical orientation — toggle with a single click
• 💾 Export your plot — download as PNG, JPG, or PDF (300 dpi)
• 📘 Built-in instructions panel — users learn the app without leaving it
• 🎨 Clean dashboard UI — powered by shinydashboard

---

🐕 Available Dog Styles

doge, doge_strong, chihuahua, eyes, gabe, glasses, tail, surprised, thisisfine, hearing, pug, ears, husky, husky_2, chilaquil

All rendered through geom_dog() from the ggdogs package.

---

📦 Requirements

• R ≥ 4.2
• Packages:

install.packages(c("shiny", "shinydashboard", "ggplot2", "dplyr", "readr", "remotes"))
remotes::install_github("R-CoderDotCom/ggdogs")

⚠️ ggdogs is not on CRAN. It must be installed from GitHub.

---

▶️ Run the App Locally

1. Clone the repository:

   git clone https://github.com/TU-USUARIO/CheemsPlot-Interactive-Dog-Styled-Boxplot-Explorer.git
   cd CheemsPlot-Interactive-Dog-Styled-Boxplot-Explorer

2. Open the project in RStudio.

3. Run the app:

   shiny::runApp()

   Or simply open app.R in RStudio and click Run App.

---

📁 Project Structure

CheemsPlot-Interactive-Dog-Styled-Boxplot-Explorer/
│
├── app.R               # Main Shiny application (UI + Server)
├── README.md           # Project documentation (this file)
├── .gitignore          # Files excluded from Git
├── data/
│    └── ejemplo.csv     
└── Images/
     └── 01_InitialMenu.png
     └── 02_outcome.png

rsconnect/ is never committed — it holds deployment credentials.

---

📘 How to Use the App

The app guides you through 8 simple steps, all available from the sidebar:

📂 Step 1 — Upload a CSV file

Use the Browse… button in the left panel to upload your .csv file. The app accepts standard comma-separated files with a header row.

📊 Step 2 — Select the Grouping variable

Choose the categorical column (e.g. breeds, groups, categories). This variable is placed on the X axis and defines the groups of the boxplot.

📈 Step 3 — Select the Numeric variable

Choose the numeric column you want to plot. This variable is placed on the Y axis and its distribution is summarized by the boxplot.

✏️ Step 4 — Type a custom Plot title

Write any title you want in the Plot title field. If you leave it empty, the app falls back to the default title "Boxplot with Dog Outliers".

🔄 Step 5 — Choose the Plot orientation

Pick between:

• Vertical — default layout, groups on the X axis
• Horizontal — flipped layout, groups on the Y axis (via coord_flip())

🐶 Step 6 — Choose the Dog style

Select which dog icon will represent the outliers. 15 styles are available: doge, doge_strong, chihuahua, eyes, gabe, glasses, tail, surprised, thisisfine, hearing, pug, ears, husky, husky_2, chilaquil.

📏 Step 7 — Adjust the Dog size

Use the slider (range 1–10) to control how big the dog icons appear on the plot. Smaller values are subtler; larger values are more expressive.

💾 Step 8 — Download your plot

Choose the Download format:

• PNG — best for presentations and web
• JPG — smaller file size
• PDF — vector format, ideal for publications

Then click Download Plot. Files are exported at 10 × 7 inches and 300 dpi via ggsave().

---

📊 What Are You Looking At?

The plot shows the distribution of your numeric variable across the groups defined by your categorical variable.

Points falling outside the interquartile range (IQR) are considered outliers and are represented with the selected dog icon instead of the default ggplot2 outlier points.

Formally, an observation is flagged as an outlier when:

value < Q1 − 1.5 · IQR    or    value > Q3 + 1.5 · IQR

Style up your analysis — statistics doesn't have to be boring. 🐕

---

📋 CSV File Requirements

Your CSV must contain at least:

• One categorical column (text) → used for grouping on the X axis
• One numeric column → used for the Y axis

Example:

breed,weight
Labrador,30
Labrador,32
Labrador,31
Labrador,75
Beagle,10
Beagle,11
Beagle,9
Beagle,45
Pug,7
Pug,8
Pug,7.5
Pug,22

---

⚠️ Notes

• No install.packages() inside app.R — installation is the user's responsibility, not the app's.
• library(ggdogs) must stay at the top level of app.R so rsconnect can detect it when building the manifest.
• The app is fully reactive: changing any input regenerates the plot instantly.
• ggsave() is used for export, so plot dimensions and dpi are consistent across formats.
• No credentials or tokens are ever committed to the repository.

---

💡 Inspiration

Inspired by the creativity of the R community — especially the idea of using custom geoms like ggdogs to bring personality into statistical graphics.

---

🧠 Why This Project?

This app demonstrates:

• Interactive data visualization with Shiny
• Advanced ggplot2 customization via custom geoms
• Reactive programming in R (reactive(), observe(), downloadHandler())
• Clean separation of UI, server logic, and reusable plotting functions
• Reproducible deployment workflow: local → GitHub → shinyapps.io / Connect Cloud

---

📸 Outcome

🖥️ Initial Dashboard View

The app opens with a clean dashboard layout: a collapsible instructions panel at the top and an empty plot area ready to render once a CSV is uploaded.

---

🐶 Dashboard in Action

Once a CSV is uploaded and the variables are selected, the boxplot is rendered with dog icons replacing the outliers. The dog style, size, orientation, and title can all be customized from the sidebar.

---

💼 Why This Matters

Outlier detection is critical in fields like:

• Finance: flagging fraudulent transactions.
• Manufacturing: detecting defective batches.
• Healthcare: identifying anomalous patient metrics.

This app makes outlier exploration accessible to non-technical users.

---

📬 Author

Ar Sullivan Santana Najera

---

⭐ If You Like It…

Give the repo a star ⭐ and share it with someone who loves dogs and data.