Skip to contents

CLAUDE.md

Source: CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Overview

ConApp is a Shiny application for mapping plant species and computing preliminary IUCN conservation status assessments following Criterion B. It is dedicated to Central Africa and built as an R package that wraps the ConR package with a web interface.

The application allows users to: - Import species occurrence data from multiple sources (SHP files, GBIF, Rainbio database, manual polygon drawing) - Map species distributions on interactive maps - Compute range parameters (EOO, AOO, number of locations) - Evaluate IUCN Red List status using Criterion B - Generate assessment reports

Commands

Running the Application

# Development mode (loads package from source)
library(pkgload)
library(shiny)
pkgload::load_all()
options(shiny.autoload.r = FALSE)
shiny::shinyApp(ui = conr_ui(), server = conr_server())

Or use the exported function:

library(ConApp)
launch()  # Launches on port 5791 by default
launch(lang = "fr")  # Launch in French

Documentation

# Build documentation from roxygen comments
devtools::document()

# View help for a function
?launch
?conr_ui

Building the pkgdown Website

# Build the pkgdown site (without running examples)
pkgdown::build_site(install = FALSE, examples = FALSE, new_process = FALSE)

# Preview the site locally
pkgdown::preview_site()

The site will be built in the docs/ directory. The site is configured to be hosted at https://umr-amap.github.io/ConApp/

Package Installation

The package depends on a remote version of ConR:

# Install from GitHub
remotes::install_github("gdauby/ConApp", upgrade = FALSE)

# ConR dependency (required)
remotes::install_github("gdauby/ConR")

Architecture

Module-Based Shiny Application

The application follows a modular Shiny architecture. Each major workflow step is isolated as a Shiny module with UI and server components:

  • Data Import Modules: Multiple sources for data input
    • module-data-full.R: Main data import orchestrator (CSV, Excel, GBIF, Rainbio)
    • module-data-poly.R: Import from SHP files and polygon drawing
    • module-data-import-gbif.R: GBIF integration
    • module-data-import-rainbio.R: Rainbio database queries
    • module-data-import-polygon.R: Manual polygon drawing on map
  • Data Processing Modules:
    • module-data-validation.R: Data validation using validation rules
    • module-data-filterout.R: Filtering out invalid records
    • module-data-variable.R: Column mapping to required variables
    • module-data-display.R: Data preview and exploration
  • Analysis Modules:
    • module-mapping.R: Interactive leaflet map for point selection and visualization
    • module-criterion_b.R: IUCN Criterion B analysis (EOO, AOO, locations computation)
    • module-data-country.R: Distribution of threatened trees in Gabon
    • module-report.R: Summary report generation

Core Application Structure

  • conr-ui.R: Main UI using bslib::page_navbar() with hidden navigation panels and a collapsible sidebar for navigation
  • conr-server.R: Main server logic coordinating all modules via reactive values (data_rv) that hold:
    • x: Main occurrence dataset
    • polygon: User-drawn or imported polygons
    • latlon: Validated latitude/longitude data

Data Flow

  1. Import: User imports data via one of the data modules → data stored in module-specific reactive
  2. Validation: Data passed through validation checks (coordinate validity, taxonomic name checks, etc.)
  3. Mapping: Validated data flows to mapping module where users can visualize and select points
  4. Analysis: Selected data analyzed using ConR package functions (EOO.computing, AOO.computing, locations.comp)
  5. Reporting: Results compiled into downloadable reports (Rmarkdown templates in inst/reports/)

Key Reactive Data Pipeline

data_full_server/data_poly_server
  → data_rv$x (main dataset)
  → mapping_server (filtered to STATUS_CONR == "IN")
  → criterion_b_server (computation of conservation parameters)
  → summary_report_server (final reports)

External Data Sources

  • GBIF: Uses rgbif package for occurrence data (gbif.R)
  • Rainbio Database: PostgreSQL database accessed via RPostgres with predefined credentials for public access (rainbio.R)
    • Connection function: conn_mydb_rb(pass = "Anyuser2022", user = "common")
    • Contains Central African plant occurrence records
  • Shapefiles: Protected areas and threat layers from PostgreSQL database (sig.R)

Internationalization

  • Supports English (default) and French via datamods::i18n()
  • Translation files in inst/i18n/fr.csv
  • Language switcher in navbar updates entire UI

Validation System

Uses the validate package with custom rules defined in validation.R: - Coordinate range checks - Empty value detection - Taxonomic name validation - Uses bdc package for biodiversity data cleaning

Theming

Custom Bootstrap theme defined in bs_theme_conr.R using bslib: - Dark theme with custom color scheme - Google font integration - Custom CSS for sidebar navigation

Important Implementation Details

  • Internal Variables: The app uses internal column naming with .__ prefix (e.g., .__latitude, .__longitude) to avoid conflicts. Always filter these out when displaying to users via unselect_internal_vars().

  • ConR Integration: Core conservation analysis functions from ConR package:

    • EOO.computing(): Extent of Occurrence (minimum convex polygon)
    • AOO.computing(): Area of Occupancy (grid-based)
    • locations.comp(): Number of threat-defined locations
    • cat_criterion_b(): IUCN category assignment

  • Leaflet Maps: All maps use base_map() helper with three basemap options (OSM, Esri, OpenTopoMap) and custom zoom controls positioned top-right.

  • Data Filtering: The STATUS_CONR column tracks record validity:

    • "IN": Valid record included in analysis
    • "OUT": Filtered out record

  • Report Templates: R Markdown templates in inst/reports/:

    • species_report.Rmd: Single species assessment
    • all_tax_report.Rmd: Multi-species summary
    • country_threat_report.Rmd: Country-level threatened species

Deployment

Deployment configuration is in deploy/app.R which loads all dependencies explicitly before launching the app. This is used for deployment to RStudio Connect or Shiny Server.