LeanClust — User Guide (Vignette)

Authors
Affiliations

Sertan Şafak

Department of Statistics, Muğla Sıtkı Koçman University, Muğla, Turkey

Muge Adnan

Department of Instructional Technology, Muğla Sıtkı Koçman University, Muğla, Turkey

Eralp Doğu

Department of Statistics, Muğla Sıtkı Koçman University, Muğla, Turkey

Serhat E. Akhanlı

Department of Statistics, Muğla Sıtkı Koçman University, Muğla, Turkey

Published

August 24, 2025

1 Overview

LeanClust is a teacher-friendly Shiny dashboard for weekly student clustering and monitoring.
This guide explains how to use the app (the methodology is covered in a companion article).

LeanClust: https://leanclust.shinyapps.io/leanclust/

LeanClust has two modes:

  • Example mode — Runs on the embedded dataset so you can explore immediately.
  • Dynamic mode — Upload your own CSV; LeanClust performs DTW + Ward.D2 clustering (k = 4) per week.

A language toggle (🇹🇷 / 🇬🇧) in the navbar switches the full UI between Turkish and English.

2 Modes

2.1 Example Mode

  • In the left sidebar, Data Source is Example Data (default).
  • Pick a Selected Week to synchronize visuals across pages.
  • Explore the pages in order: Cluster Table → Transitions → Cluster Summaries → Developers.

2.2 Dynamic Mode

  • In the left sidebar choose Dynamic Data, upload your CSV, then click Analyze.
  • Column prefix conventions (case-insensitive):
    • id* — Student identifier (exactly one column required)
    • wk* — Week label (exactly one column required)
    • at* — Attendance/activity/engagement features (0+ columns)
    • sc* — Score/assessment features (quiz, midterm, final, etc.; 0+ columns)

Minimal CSV example:

id_student,wk_label,at_reading,at_practice,sc_quiz,sc_midterm S001,Week 1,0.8,0.6,, S001,Week 2,0.7,0.5,65, S001,Week 5,0.6,0.7,,72 S002,Week 1,0.4,0.3,, S002,Week 2,0.5,0.4,55, S002,Week 9,0.6,0.5,,60 —

3 UI Tour

LeanClust uses a top navbar with four pages and a persistent left sidebar.

3.2 Page 1 — Cluster Table

What you see on app start
A wide table with Student ID rows and Week columns. Each cell is a colored band representing the cluster for that week.

Controls and actions:

- ① Search box — type a Student ID to filter rows (partial text supported).
- ② Week slider — selects the last highlighted week; earlier weeks remain color-filled.
- ③ Legend — fixed cluster names/colors: Reserved, Peripheral, Engaged, Strategic.

How to use:

1. Enter a Student ID in ① to focus on a learner.
2. Move ② to change the last highlighted week.

3.3 Page 2 — Transitions (Sankey)

Two Sankey diagrams visualize cluster transitions:

  • ① Left diagram: From Week → (To Week − 1).
  • ② Right diagram: From Week → To Week.
  • Colors — nodes/links match the four cluster colors used across the app.

How to use:

1. Set From Week and To Week in the sidebar.
2. Compare ① (pre-final) with ② (final).

If To Week is the first week, ① will be empty by design.

3.4 Page 3 — Cluster Summaries

Per-cluster summary cards for the Selected Week:

  • ① Variables to Display — choose metrics (labels are prettified).
  • ② Card header — cluster name (color-coded).
  • ③ Count badge — number of students in the cluster.
  • ④ Metric rows — mean values of selected variables.

Changing Selected Week updates all cards.

3.5 Page 4 — Developers

This page lists the developers of the application, their contact details, and provides a link to the GitHub repository. You can easily access the GitHub repository by clicking the button.

4 Cluster Labels & Colors

LeanClust consistently uses the following labels and colors across all visuals:

  1. Reserved Performers — #0072B2
  2. Peripheral Passers — #D55E00
  3. Engaged Achievers — #F0E442
  4. Strategic Achievers — #009E73

5 Quick Start Checklist

  1. Open LeanClust
  2. Example mode (default): pick a Selected Week and explore pages 1→4.
  3. Dynamic mode: upload a CSV following id* / wk* / at* / sc* prefixes and click Analyze.
  4. Use From/To Week to inspect Transitions; adjust Selected Week for Summaries.
  5. Toggle 🇹🇷 / 🇬🇧 for instant bilingual UI.