Shiny Performance Tracking

This repository provides a complete pipeline for processing rodent training data, converting raw .mat files into structured .rds files, extracting behavioral metrics, and visualizing performance trends in an interactive Shiny dashboard.

Overview

The system enables researchers to:

• Import experimental data from BControl or Bpod sessions.
• Process and transform data into a clean, analysis-ready format.
• Automatically aggregate session-level metrics into TRAINING.csv.
• Explore data interactively with Shiny visualizations.

Repository Structure

shiny-performance-tracking/
├── docs/
│   ├── architecture.md
│   ├── data_dictionary.md
│   ├── usage_guide.md
│   └── setup_notes.md
├── shiny_app/
│   ├── app.R
│   ├── TRAINING.csv         # Main dataset for the dashboard
│   ├── full_TRAINING.csv    # Optional, long-term aggregation
│   └── functions/           # Modular ggplot-based visualization scripts
├── utility_functions/       # File parsing and processing logic
│   ├── ConvertToRDS.R
│   ├── ReadBcontrolData.R
│   ├── ReadBpodData.R
│   ├── ReadData.R
│   ├── ReadTrialData.R
│   └── TRAININGtoCSV.R
├── .Rprofile                # R environment configuration and package management
├── ExtractSaveData.R        # Main ETL pipeline: from .mat to .csv
└── README.md                # Overview + links to docs

Shiny App

Launch the app locally:

setwd("shiny_app")
shiny::runApp()

The dashboard includes:

• Stage tracking by animal
• Correct ratio trends
• Trial completion rates
• Choice direction distributions

Dependencies

The system uses several R packages organized by functionality:

Core Data Science

• tidyverse - Data manipulation and visualization
• magrittr - Pipe operator for cleaner code
• parallel - Parallel computing support

Data Import/Export

• R.matlab - MATLAB file support
• readr - Fast data import
• tibble - Modern data frames

Data Manipulation

• stringr - String manipulation
• forcats - Categorical data handling
• purrr - Functional programming
• zoo, chron, padr - Time series analysis

Visualization

• ggplot2 - Base plotting
• ggpubr - Enhanced ggplot2
• ggrepel - Label positioning
• plotly - Interactive plots
• gridExtra - Multi-plot layouts

Shiny and Web

• shiny - Web application framework
• shinyjs - JavaScript integration
• DT - Interactive tables
• kableExtra - Enhanced tables
• rmarkdown, knitr - Dynamic documents

All dependencies are automatically loaded through .Rprofile, which also:

• Manages utility function loading using a vector-based approach
• Configures system-specific data paths
• Sets up the R environment for optimal performance

Documentation

• Project Architecture
• Data Dictionary
• Usage Guide
• Server Setup Notes

Notes

• ExtractSaveData.R supports batch processing and parallel conversion.
• New .mat files added to the source directory are automatically processed and appended.
• The system supports both session-level and trial-by-trial data (optional).
• Utility functions are loaded dynamically through .Rprofile for better maintainability.