Changelog

nemeton 0.14.1

Date: 2026-02-18

UI Improvements

Made the recent projects section collapsible using the same Bootstrap 5 collapse pattern as the commune search and project form sections

Bug Fixes

Fixed namespace issues in i18n and energy indicator tests
Fixed mock bindings for lookup_species_threshold using unlockBinding
Suppressed expected OSM tile download warnings in export tests
Fixed various test stability issues (memory, timeouts, namespace prefixes)

Documentation

Updated README with app screenshot and badge updates
Prepared package for CRAN submission

nemeton 0.14.0

Date: 2026-02-10

Test Suite Stabilization

Bug Fixes

Fixed ExtendedTask global state leak in mod_home retry test that blocked mod_project testServer calls when running the full test suite
- Mock promises::future_promise to prevent multisession worker spawning
- Return terminal state from read_progress_state to stop later::later polling loop
- Restore future::plan on exit via withr::defer
Suppressed expected warnings in test-workflow, test-visualization, test-mod_map, and test-mod_synthesis
Renamed test files with z/zz prefix for stable execution ordering

Documentation

Added Mistral API key example in nemetonApp guide vignette

Tests

All 9272 tests passing (0 warnings)
R CMD check: 0 errors, 0 warnings

nemeton 0.13.0

Date: 2026-02-08

CI/CD Optimization

Optimized CI workflow with timeout, split check and coverage jobs
Suppressed expected warnings in test suite (normalize, locale patterns)
Fixed French locale support in match.arg error patterns

nemeton 0.12.0

Date: 2026-02-05

Phase 9 Finalization - MVP 0.7.0 Complete

New Features

PDF Report Generation (generate_report_pdf())
- Quarto-based reports with professional layout
- Fallback to base R graphics when Quarto unavailable
- Automatic Quarto installation via ensure_quarto_installed()
- Bilingual support (French/English)
GeoPackage Export (export_geopackage())
- Export family scores with geometry for GIS analysis
- Full spatial data preservation
nemetonApp Synthesis Tab
- AI-generated analysis with expert profiles
- Integrated comment editor
- Real-time PDF generation with progress indicator

Documentation

New vignette: “Guide de l’Application nemetonApp”
Updated README with nemetonApp section
Enhanced pkgdown reference for Shiny functions

Bug Fixes

Fixed TWI normalization windows for F2 soil fertility ([2.5, 10] range)
Fixed R3 drought risk raster extent mismatch
Fixed non-ASCII characters in service_export.R
Added data.table to Suggests for fasterRaster compatibility

Tests

All 3447 tests passing
R CMD check: 0 errors, 0 warnings, 2 notes

nemeton 0.8.0

Date: 2026-01-25

New Features

nemetonApp - Interactive Shiny Application

run_app() - Launch the nemetonApp Shiny application
- Interactive parcel selection on a map (Leaflet)
- French commune search with autocomplete
- Calculate all 31 nemeton indicators automatically
- Visualize results with 12-family radar charts
- Export PDF reports and GeoPackage data
- Bilingual support (French/English)
Application Architecture
- app_ui.R - bslib-based responsive UI with Bootstrap 5
- app_server.R - Modular server with reactive state management
- app_config.R - Configuration constants and indicator families
- utils_theme.R - WCAG 2.1 AA accessible theme
- utils_i18n.R - Internationalization with 200+ messages
Accessibility (WCAG 2.1 AA)
- Color contrast ratio >= 4.5:1 for text
- Colorblind-friendly viridis palettes
- Minimum touch target 44×44px
- Focus visible indicators
- Keyboard navigation support
Data Services
- service_communes.R - French commune search API
- service_cadastre.R - Cadastral parcel retrieval
- service_project.R - Project management and persistence

Bug Fixes

Fixed \dontrun missing brace in service_communes.R documentation
Fixed integer type for symbol_shapes in accessibility config
Updated indicator count test (29 → 31 indicators)

Dependencies

Added shiny (>= 1.8.0) to Imports
Added bslib (>= 0.6.0) to Imports
Added htmltools (>= 0.5.7) to Imports
Added leaflet (>= 2.1.0) to Suggests
Added cicerone (>= 1.0.0) to Suggests (guided tour)
Added shinyWidgets (>= 0.8.0) to Suggests
Added rappdirs to Suggests (project directories)

nemeton 0.6.2

Date: 2026-01-24

Changes

Data consolidation: Merged massif_demo_units and massif_demo_units_extended into a single dataset with 89 columns (29 indicators, 12 family composites, normalized values)
Tests: Fixed 19 skipped tests, now 1478 tests passing (0 skipped)
Documentation: Simplified README from 846 to 138 lines
Fixtures: Added synthetic cadastral file for integration tests

nemeton 0.6.1

Date: 2026-01-23

Changes

Fix pkgdown references to obsolete v0.1.0 indicators
Add lasR remote for GitHub Actions CI

nemeton 0.6.0 (Development)

v0.6.0 - Legacy Indicators Removal

Date: 2026-01-23

BREAKING CHANGES ⚠️

Removed Legacy Indicators (v0.1.0)

The original 5 MVP indicators have been removed in favor of the comprehensive 12-family framework (32+ indicators). This is a breaking change for code using v0.1.0 indicators.

Removed Functions

indicator_carbon() - Use instead: indicator_carbon_biomass() (C1) or indicator_carbon_ndvi() (C2)
indicator_biodiversity() - Use instead: indicator_biodiversity_protection() (B1), indicator_biodiversity_structure() (B2), or indicator_biodiversity_connectivity() (B3)
indicator_water() - Use instead: indicator_water_network() (W1), indicator_water_wetlands() (W2), or indicator_water_twi() (W3)
indicator_fragmentation() - Use instead: indicator_landscape_fragmentation() (L1) or indicator_landscape_edge() (L2)
indicator_accessibility() - Use instead: indicator_social_accessibility() (S2) or indicator_social_trails() (S1)

Migration Guide

Before (v0.1.0-v0.5.x):

# Old API
results <- nemeton_compute(
  units, layers,
  indicators = c("carbon", "biodiversity", "water")
)

After (v0.6.0+):

# New API with family-based indicators
results <- nemeton_compute(
  units, layers,
  indicators = c("carbon_biomass", "biodiversity_protection", "water_twi")
)

# Or use list_indicators() to see all available indicators
available <- list_indicators(return_type = "details")

Updated Core Functions

nemeton_compute() - Now uses list_indicators() for available indicators
list_indicators() - Returns all 31 indicators from the 12-family framework
compute_indicator() - Dynamic dispatch supporting all family-based indicators

Files Removed

R/indicators-biophysical.R - Legacy indicator implementations (567 lines)
tests/testthat/test-indicators-biophysical.R - Legacy tests (414 lines, 26 tests)

Rationale

The legacy indicators were functional placeholders from the v0.1.0 MVP. The new 12-family framework (introduced in v0.2.0-v0.4.0) provides:

More comprehensive coverage: 31 indicators vs 5
Better scientific foundation: Species-specific allometric models, multiple data sources
Clearer organization: 12 families (C, W, F, L, B, R, T, A, S, P, E, N)
Improved flexibility: Multiple sub-indicators per ecosystem service

All legacy indicators had superior replacements available since v0.2.0 (January 2026).

nemeton 0.5.2

v0.5.2 - Tutorial 09 Sampling + TSP

Date: 2026-01-23

New Features

Tutorial 09: Échantillonnage de calibration LiDAR HD + TSP (180 min)

Dimensionnement optimal - Calcul du nombre de placettes basé sur la formule n = t² × CV² / E²
- Fonctions calculate_sample_size() et sample_size_table()
- Tableau de référence interactif pour CV (10-40%) et erreur (5-20%)
- Correction pour population finie
Sampling Frame - Construction d’une grille de candidats avec contraintes terrain
- Filtrage par couvert forestier (≥70%) et pente (≤45%)
- Utilisation des données T01/T03/T07 (zone_etude, bd_foret, mnt, chm_complet)
Stratification triple - Basée sur 3 critères forestiers
- Hauteur CHM LiDAR : 4 classes (H1-H4) par quartiles
- Type de peuplement (BD Forêt v2 tfv) : FEU/CON/MIX/POP/AUT
- Position topographique (TPI) : Bas/Milieu/Haut de pente
- Calcul TPI avec focal() (rayon 100m)
Tirage GRTS stratifié - Échantillonnage spatialement équilibré
- Package spsurvey::grts() avec allocation proportionnelle
- Oversample par strate pour placettes de remplacement
- Fallback BalancedSampling::lpm2 si GRTS échoue
Réseau de chemins - Construction réseau avec sfnetworks depuis BD TOPO
- Filtrage chemins praticables à pied
- Calcul poids avec edge_length()
Optimisation TSP - Parcours optimal avec package TSP
- Méthode nearest insertion + 2-opt
- Visualisation avec distinction Base/Remplacement
Export terrain - Formats multiples pour GPS
- GeoPackage (SIG)
- GPX (navigation GPS)
- CSV (tableau récapitulatif avec coordonnées WGS84)

Improvements

Harmonisation data_dir - Chemin unifié sur tous les tutoriels T01-T09
- ~/.local/share/nemeton/tutorial_data
- Suppression variable cache_dir dans T08

Dependencies

Added spsurvey (>= 5.0.0) to Suggests
Added BalancedSampling (>= 1.6.0) to Suggests
Added sfnetworks (>= 0.6.0) to Suggests
Added TSP (>= 1.2.0) to Suggests
Added tidygraph (>= 1.2.0) to Suggests
Added igraph (>= 1.4.0) to Suggests

Documentation

Updated vignettes/tutorial-guide.Rmd with Tutorial 09
Updated TUTORIAL_INSTALL.md with Tutorial 09

References: - Stevens, D. L., & Olsen, A. R. (2004). Spatially balanced sampling of natural resources. JASA, 99(465), 262-278. - Grafström, A., & Tillé, Y. (2013). Doubly balanced spatial sampling with spreading and restitution of auxiliary totals. Environmetrics, 24(2), 120-131. - Hahsler, M., & Hornik, K. (2007). TSP—Infrastructure for the traveling salesperson problem. Journal of Statistical Software, 23(2).

nemeton 0.5.1

v0.5.1 - Tutorial 08 Coregistration

Date: 2025-01-18

New Features

Tutorial 08: Coregistration LiDAR/Terrain (130 min)

Problématique GPS - Précision GPS sous couvert forestier (2-10 m)
Corrélation MNH/Terrain - Recalage par corrélation croisée
lidaRtRee::coregistration() - Recherche translation optimale (dx, dy)
Traitement parallèle - future_lapply() pour lots de placettes
Analyse statistique - Tests de significativité, visualisation vecteurs
Export - CSV et GeoPackage pour utilisation SIG

Documentation

Updated vignettes/tutorial-guide.Rmd with Tutorial 08
Updated TUTORIAL_INSTALL.md with Tutorial 08

Reference: Monnet, J.-M., & Mermin, É. (2014). Cross-correlation of diameter measures for the co-registration of forest inventory plots with airborne laser scanning data. Forests, 5(9), 2307-2326.

nemeton 0.5.0

v0.5.0 - Tutorial 07 & CRAN Compliance

Date: 2025-01-18

Overview

Release featuring the complete Tutorial 07 (Advanced LiDAR) and CRAN compliance improvements. All 7 interactive tutorials are now complete (195/195 tasks).

New Features

Tutorial 07: LiDAR Avancé (90 min)

LAScatalog Management - Multi-tile LiDAR processing with lidR
lasR Pipelines - Ultra-fast C++ processing for DTM/CHM generation
Individual Tree Detection (ITD) - Tree segmentation with lidaRtRee
Gap & Edge Detection - Forest structure analysis
Area-Based Approach (ABA) - Model calibration and wall-to-wall prediction
BABA Exploration - Rapid LiDAR metrics without field calibration
Parallelization - future_lapply() for tile-based processing
Incremental Caching - Resume interrupted processing
OSO Forest Mask - Land cover filtering for predictions

Dependencies

Added lasR to Suggests (from r-lidar.r-universe.dev)
Added lidaRtRee to Suggests

Documentation

Updated vignettes/tutorial-guide.Rmd with Tutorial 07
Updated TUTORIAL_INSTALL.md with lasR/lidaRtRee installation
Updated quickstart guide with Tutorial 07 instructions

CRAN Compliance

Removed development artifacts (RELEASE_*.md, .RData, .Rhistory, etc.)
Updated .Rbuildignore and .gitignore
Excluded spec-kit directories from version control

nemeton 0.4.0

v0.4.0 - Complete 12-Family Ecosystem Services Referential

Date: 2026-01-05

Overview

Major release completing the 12-family ecosystem services referential with 4 new indicator families (S, P, E, N) and advanced multi-criteria analysis tools. This release adds 11 new indicator functions, 3 analysis functions, and brings the total to 29 indicators across 12 families.

New Indicator Families

indicator_social_trails() (S1) - Trail density
- Calculates recreational trail density (km/ha) from OSM or local data
- Supports footways, cycleways, and bridleways
- Output: 0-5+ km/ha trail density
indicator_social_accessibility() (S2) - Multimodal accessibility score
- Distance-based scoring for road, parking, and public transport access
- Configurable distance thresholds and weights
- Output: 0-100 accessibility score
indicator_social_proximity() (S3) - Population proximity
- Population within configurable buffer zones (5/10/20 km)
- Supports INSEE population grid or custom data
- Output: Total population count within buffers

Productive & Economic Family (Famille P) - 3 Indicators

indicator_productive_volume() (P1) - Standing timber volume
- IFN-based allometric equations by species
- Genus-level fallback for rare species
- Output: m³/ha standing volume
indicator_productive_station() (P2) - Site productivity index
- Fertility × climate × species interaction
- Based on French forestry station classification
- Output: m³/ha/yr potential productivity
indicator_productive_quality() (P3) - Timber quality score
- Form factor, diameter distribution, defect assessment
- Configurable quality criteria weights
- Output: 0-100 quality score

Energy & Climate Family (Famille E) - 2 Indicators

indicator_energy_fuelwood() (E1) - Fuelwood potential
- Harvest residues + coppice biomass estimation
- Species-specific conversion factors
- Output: tonnes DM/ha/yr mobilizable fuelwood
indicator_energy_avoidance() (E2) - CO2 emission avoidance
- ADEME emission factors for energy substitution
- Supports energy and material substitution scenarios
- Output: tCO2eq/ha/yr avoided emissions

Naturalness & Wilderness Family (Famille N) - 3 Indicators

indicator_naturalness_distance() (N1) - Infrastructure distance
- Distance to roads, buildings, powerlines from OSM
- Minimum distance to nearest infrastructure
- Output: meters to nearest infrastructure
indicator_naturalness_continuity() (N2) - Forest patch continuity
- Connected forest area calculation
- Based on landscape patch analysis
- Output: hectares of continuous forest
indicator_naturalness_composite() (N3) - Wilderness composite index
- Multiplicative aggregation of N1 × N2 × T1 × B1
- Weighted aggregation option available
- Output: 0-100 wilderness score

New Analysis Functions

Multi-Criteria Decision Support

identify_pareto_optimal() - Pareto optimality analysis
- Identifies non-dominated solutions across multiple objectives
- Supports both maximization and minimization objectives
- Returns data with is_optimal column for Pareto-optimal parcels
cluster_parcels() - Multi-family clustering
- K-means and hierarchical clustering methods
- Automatic optimal k determination via silhouette analysis
- Returns cluster assignments with centroid profiles
plot_tradeoff() - Trade-off visualization
- 2D scatterplot with optional Pareto frontier overlay
- Color and size mapping for additional dimensions
- Label support for parcel identification

Enhanced Features

12-axis radar plots - nemeton_radar() now supports all 12 families
12×12 correlation matrix - compute_family_correlations() extended
12-family hotspot detection - identify_hotspots() updated
Normalization presets - Added for S, P, E, N families

New Demo Dataset

massif_demo_units_extended - Complete 12-family reference dataset
- 20 demo parcels with all 29 indicators
- 12 pre-calculated family composite indices
- Synthetic but realistic value distributions

New Vignettes

complete-referential_fr.Rmd - 12-family workflow demonstration
multi-criteria-optimization_fr.Rmd - Pareto, clustering, and trade-off analysis

Dependencies

Added cluster package dependency for silhouette analysis
Added ggrepel to Suggests for label positioning

Documentation

Updated README with v0.4.0 feature highlights
Updated pkgdown site configuration
Full roxygen2 documentation for all new functions

nemeton 0.3.0 (Development)

v0.3.0 MVP - Multi-Family Extension (B, R, T, A)

Status: ✅ v0.3.0 Complete (845+ tests passing, 100% backward compatible)

Overview

Extension of the ecosystem service indicator framework with 4 new families (B, R, T, A), bringing the total to 9 of 12 planned families implemented. This release adds 10 new indicator functions and enhances the family aggregation and visualization system.

New Indicator Families

Biodiversity Family (Famille B) - 3 Indicators

indicator_biodiversity_protection() (B1) - Protected area coverage
- Calculates percentage of forest parcel within protected zones (ZNIEFF, Natura2000, etc.)
- Supports local or remote protected area datasets
- Output: 0-100% protection coverage
- Spatial overlay with area-weighted calculation
indicator_biodiversity_structure() (B2) - Structural diversity index
- Shannon diversity index across vegetation strata, age classes, and species
- Configurable weights for each diversity component (default: strata 0.4, age 0.3, species 0.3)
- Optional height coefficient of variation (CV) integration
- Output: 0-100 diversity score
- Handles monoculture scenarios (low diversity → low scores)
indicator_biodiversity_connectivity() (B3) - Ecological connectivity
- Distance to ecological corridors (TVB - Trame Verte et Bleue)
- Supports edge and centroid distance methods
- Configurable max distance threshold (default: 5000m)
- Fallback scoring when corridor data unavailable
- Output: Distance in meters (lower = better connectivity)

Risk & Resilience Family (Famille R) - 3 Indicators

indicator_risk_fire() (R1) - Fire risk index
- Multi-factor fire susceptibility: slope + species + climate
- Species flammability coefficients (conifer > broadleaf)
- Slope amplification (higher slope = faster fire spread)
- Optional climate data integration (temperature, precipitation)
- Output: 0-100 risk score (higher = more vulnerable)
indicator_risk_storm() (R2) - Storm vulnerability index
- Wind damage risk: tree height × stand density × exposure
- Height coefficient (taller trees more vulnerable)
- Density factor (dense stands = higher windthrow risk)
- Topographic exposure from DEM (exposed ridges = higher risk)
- Output: 0-100 vulnerability score
indicator_risk_drought() (R3) - Drought stress index
- Combines water availability (TWI) and species drought tolerance
- Species tolerance coefficients (drought-resistant vs. water-demanding)
- Optional precipitation data integration
- Low TWI + intolerant species = high stress
- Output: 0-100 stress score

Temporal Dynamics Family (Famille T) - 2 Indicators

indicator_temporal_age() (T1) - Stand age/ancientness
- Historical forest age from BD Forêt or Cassini maps
- Ancient forest detection (age > 150 years)
- Supports multi-source age estimation
- Output: Years since establishment (or age class)
- Handles missing historical data gracefully
indicator_temporal_change() (T2) - Land cover change rate
- Temporal change detection using Corine Land Cover multi-dates
- Calculates change rate between two periods
- Supports custom date ranges
- Identifies stable vs. dynamic forests
- Output: % change per year (or absolute area change)
- Leverages existing nemeton_temporal() infrastructure

Air Quality & Microclimate Family (Famille A) - 2 Indicators

indicator_air_coverage() (A1) - Tree canopy coverage
- Percentage of tree cover within 1km buffer
- High-resolution vegetation data (sentinel-2 or lidar-derived)
- Urban microclimate regulation potential
- Output: 0-100% coverage in buffer zone
- Supports custom buffer distances
indicator_air_quality() (A2) - Air quality index
- Integration with ATMO air quality data (when available)
- Fallback: distance to pollution sources (roads, industry)
- Supports custom air quality datasets
- Output: 0-100 air quality score (higher = better)
- Proxy mode for areas without monitoring stations

Extended Functions

create_family_index() - New “min” aggregation method
- Added conservative worst-case aggregation: method = "min"
- Useful for risk assessment (score = worst sub-indicator)
- Joins existing methods: mean, weighted, geometric, harmonic
- Example: create_family_index(data, method = "min")
nemeton_radar() - Comparison mode for multiple units
- New: compare multiple forest parcels on same radar chart
- Overlaid polygons for visual comparison
- Syntax: nemeton_radar(data, unit_id = c(1, 5, 10), mode = "family")
- Supports 9-12 axes dynamically (adapts to available families)
- Enhanced legend and color differentiation

Testing

186 new tests for v0.3.0 families
- Biodiversity (B1-B3): 56 tests (protection zones, diversity indices, corridors)
- Risk (R1-R3): 52 tests (fire models, storm factors, drought stress)
- Temporal (T1-T2): 38 tests (historical data, change detection)
- Air (A1-A2): 28 tests (coverage buffers, quality indices)
- Integration: 12 tests (multi-family workflows, normalization, radar)
Total test suite: 845+ tests passing (up from 659)
100% backward compatibility verified with v0.2.0 workflows

Use Cases

Conservation prioritization: Identify high biodiversity + low risk parcels
Climate adaptation planning: Map drought stress + storm vulnerability
Urban forestry: Quantify air quality and microclimate benefits
Historical ecology: Detect ancient forests + track land use change
Multi-criteria decision support: 9-family composite indices for holistic management

Implemented Families Status (9/12)

✅ C - Carbon & Vitality (C1-C2)
✅ B - Biodiversity (B1-B3) - NEW v0.3.0
✅ W - Water Regulation (W1-W3)
✅ A - Air Quality & Microclimate (A1-A2) - NEW v0.3.0
✅ F - Soil Fertility (F1-F2)
✅ L - Landscape & Aesthetics (L1-L2)
✅ T - Temporal Dynamics & Trame (T1-T2) - NEW v0.3.0
✅ R - Risk Management & Resilience (R1-R3) - NEW v0.3.0
⏳ S - Social & Recreational (planned v0.4.0)
⏳ P - Productive & Economic (planned v0.4.0)
⏳ E - Energy & Climate (planned v0.4.0)
⏳ N - Naturalness & Night (planned v0.4.0)

nemeton 0.2.0 (Development)

v0.2.0 - Phase 9: Multi-Family System (US6)

Status: ✅ Phase 9 Complete (659 tests passing, +46 from Phase 8)

New Functions

Multi-Family Aggregation & Visualization

create_family_index() - Family-level composite scores
- Aggregates sub-indicators into family indices (family_C, family_W, etc.)
- Automatic detection of family prefixes (C1, C2 → family_C)
- 4 aggregation methods: mean, weighted, geometric, harmonic
- Custom weights per family
- Supports all 12 families (C, B, W, A, F, L, T, R, S, P, E, N)
- Returns sf object with added family_* columns

Extended Functions

normalize_indicators() family support
- Added by_family parameter for family-aware workflows
- Auto-detection of family indicators (C1, W1, F1 pattern)
- Backward compatible with v0.1.0 indicators (carbon, water, etc.)
- When by_family = TRUE: normalizes in-place (suffix = ““, keep_original = FALSE)
nemeton_radar() multi-family mode
- New mode parameter: “indicator” (default) or “family”
- Family mode: displays 4-12 family axes dynamically
- Auto-detects family_* columns when mode = “family”
- Backward compatible with indicator mode
- Enhanced unit_id handling: supports both ID matching and numeric row indices

Helper Functions (Internal)

detect_indicator_family() - Extract family code from indicator name
get_family_name() - Full family name from code (bilingual FR/EN)

Testing

46 new tests for multi-family system
- create_family_index(): 9 tests (aggregation methods, weights, NA handling)
- normalize_indicators() family support: 3 tests (auto-detection, by_family mode)
- nemeton_radar() family mode: 4 tests (multi-family display, validation)
- Integration: 5 tests (end-to-end workflows, temporal integration)
- Family detection: 2 tests (all 12 families)
Total test suite: 659 tests passing (up from 613)
2 minor test issues: plot data structure check, locale-dependent error message
Full backward compatibility maintained

Technical Details

Family Detection: Regex pattern ^[A-Z][0-9] matches C1, W1, F1, etc.
Aggregation Methods:
- Mean/Weighted: Handles NA values with weight renormalization
- Geometric: exp(mean(log(values))) with negative value handling
- Harmonic: n / sum(1/x) with zero value handling
12 Family Codes:
- C (Carbon & Vitality), B (Biodiversity), W (Water Regulation)
- A (Air Quality & Microclimate), F (Soil Fertility), L (Landscape & Aesthetics)
- T (Temporal Dynamics), R (Risk Management), S (Social & Recreational)
- P (Productive & Economic), E (Energy & Climate), N (Naturalness)

Use Cases

Multi-dimensional assessment: Compare ecosystem services across 12 families
Custom weighting: Priority to specific families (e.g., 60% carbon, 40% water)
Radar visualization: Visual profiling of forest parcels across all families
Family-level reporting: Aggregate detailed indicators into comprehensible family scores

v0.2.0 - Phase 8: Infrastructure Multi-Temporelle (US1)

Status: ✅ Phase 8 Complete (613 tests passing)

New Functions

Temporal Analysis Framework - 2 Core Functions + 2 Visualizations

nemeton_temporal() - Multi-period temporal dataset creation
- Combines nemeton_units objects from different time periods
- Automatic unit alignment tracking across periods
- Support for ISO dates and custom period labels
- Metadata: dates, period labels, alignment status
- Returns nemeton_temporal S3 class with print/summary methods
calculate_change_rate() - Temporal change rate calculation
- Computes absolute change rates (units per year)
- Computes relative change rates (% per year)
- Supports indicator selection or “all” indicators
- Configurable start/end periods
- Handles NA values appropriately
- Returns sf object with _rate_abs and _rate_rel columns
plot_temporal_trend() - Time-series line plots
- Line plots showing indicator evolution over time
- Single indicator: all units on one plot
- Multiple indicators: faceted plots (2 columns)
- Optional mean trend line overlay
- Unit selection support
- Automatic date handling from temporal metadata
plot_temporal_heatmap() - Indicator evolution heatmap
- Heatmap showing all indicators across periods for one unit
- Optional normalization to 0-100 scale
- Viridis color scale
- Value labels on tiles
- Indicator selection support
- Useful for single-unit profiling

S3 Methods

print.nemeton_temporal() - Console summary
- Shows number of periods and units
- Date range if available
- Warns about incomplete alignment
- Lists available indicators
summary.nemeton_temporal() - Detailed statistics
- Per-period summaries (unit counts, indicator ranges)
- Mean values for each indicator per period
- Alignment information

Technical Details

Temporal Framework:
- Unit ID tracking with configurable column (default: “parcel_id”)
- Automatic alignment detection (units present in all periods vs. incomplete)
- Flexible date handling (ISO dates, years, or custom labels)
- Preserves all sf attributes and geometry
Change Rates:
- Time difference calculation from dates or period names
- Absolute rate: (value_end - value_start) / years
- Relative rate: ((value_end / value_start) - 1) * 100 / years
- NA propagation for missing data
Visualizations:
- ggplot2-based with theme_minimal
- Date axis with automatic formatting
- Faceting for multiple indicators
- Viridis colormap for heatmaps
- Responsive layouts (legend positions, text angles)

Testing

79 new tests for temporal infrastructure
- nemeton_temporal(): 13 tests (creation, alignment, validation)
- calculate_change_rate(): 13 tests (absolute/relative rates, NA handling)
- print/summary methods: 3 tests (output format)
- plot_temporal_trend(): 11 tests (single/multiple indicators, unit selection)
- plot_temporal_heatmap(): 10 tests (normalization, indicator selection)
- Integration: 4 tests (multi-period workflows, 3+ periods)
Total test suite: 613 tests passing (up from 584)
Full backward compatibility maintained

Use Cases

Longitudinal monitoring: Track indicator evolution over 5-10+ years
Management impact: Compare before/after forest intervention
Climate change: Detect long-term trends in carbon stock, water regulation
Scenario comparison: Visualize different management trajectories

v0.2.0 - Phase 7: Famille L (Landscape/Paysage)

Status: ✅ Phase 7 Complete (584 tests passing)

New Indicator Functions

Landscape Family (Famille L) - 2 Indicators

indicator_landscape_fragmentation() (L1) - Forest fragmentation metric
- Counts number of forest patches within a buffer zone around each parcel
- Uses connected component labeling (terra::patches with 8-neighbor connectivity)
- Configurable buffer distance (default: 1000m)
- Configurable forest definition via landcover codes
- Output: Number of discrete forest patches (higher = more fragmented)
- Zero buffer option for parcel-only analysis
indicator_landscape_edge() (L2) - Edge-to-area ratio
- Calculates perimeter-to-area ratio for forest parcels
- Formula: Edge density = perimeter (m) / area (ha)
- Higher values indicate greater edge effect and fragmentation
- Output: m/ha (meters of edge per hectare)
- Uses sf geometry operations for precise boundary calculations

Technical Details

L1 Fragmentation:
- Buffer zone creation using sf::st_buffer()
- Landcover cropping and masking with terra
- Forest mask creation using terra::app() with custom classification
- Connected component analysis: terra::patches(directions = 8)
- Handles zero-forest scenarios gracefully
L2 Edge Density:
- Boundary extraction: sf::st_cast() to MULTILINESTRING
- Perimeter calculation: sf::st_length()
- Area calculation: sf::st_area() converted to hectares
- No dependencies on raster layers (geometry-only)

Testing

49 new tests for landscape family indicators
- L1 fragmentation: 13 tests (patch counting, buffer effects, forest definitions)
- L2 edge: 11 tests (geometry scaling, parcel size effects, validation)
- Integration: 8 tests (combined workflow, dataframe integration, correlation analysis)
- Edge cases: 5 tests (empty units, single parcels, full dataset)
Total test suite: 584 tests passing (up from 535)
Full backward compatibility maintained

v0.2.0 - Phase 6: Famille F (Fertilité des Sols)

Status: ✅ Phase 6 Complete (535 tests passing)

New Indicator Functions

Soil Family (Famille F) - 2 Indicators

indicator_soil_fertility() (F1) - Soil fertility classification
- Extracts fertility scores from soil data (raster or vector)
- Supports BD Sol (French soil database) or equivalent pedological data
- Output: 0-100 scale (higher = more fertile)
- Auto-normalizes input values to 0-100 range
- Supports both raster and vector soil layers (with area-weighted averaging)
indicator_soil_erosion() (F2) - Erosion risk index
- Calculates erosion risk by combining slope and land cover protection
- Formula: Risk = slope × (1 - forest_protection)
- High slope + low forest cover = high erosion risk
- Output: 0-100 risk score
- Uses terra for slope calculation and land cover analysis

Internal Utilities

Soil Data Extraction
- extract_fertility_from_raster() - Raster-based fertility extraction
- extract_fertility_from_vector() - Vector-based fertility with spatial join
- Area-weighted averaging for overlapping soil polygons
- Automatic CRS harmonization

Testing

37 new tests for soil family indicators
- F1 fertility: 11 tests (raster/vector extraction, normalization, error handling)
- F2 erosion: 17 tests (slope-cover combination, forest definitions, edge cases)
- Integration: 9 tests (combined workflow, correlation analysis, dataframe integration)
- 1 skipped test (vector soil data - future enhancement)
Total test suite: 535 tests passing (up from 498)
Full backward compatibility maintained

Technical Details

F1 Fertility:
- Flexible input: accepts any raster or vector soil layer
- Linear normalization: (value - min) / (max - min) × 100
- Vector mode: area-weighted spatial join with soil polygons
F2 Erosion:
- Slope from DEM using terra::terrain(v="slope")
- Forest mask using terra::app() for multi-value classification
- Protection factor: 1 = full forest, 0 = no forest
- Normalized to 0-100 scale (max slope = 90°)

v0.2.0 - Phase 5: Famille W (Eau/Infiltrée)

Status: ✅ Phase 5 Complete (498 tests passing)

New Indicator Functions

Water Family (Famille W) - 3 Indicators

indicator_water_network() (W1) - Hydrographic network density
- Calculates stream/river network length density within or near forest parcels
- Supports buffer distance parameter for proximity analysis
- Output: km/ha (kilometers of watercourse per hectare)
- Higher values = greater hydrological connectivity
indicator_water_wetlands() (W2) - Wetland coverage percentage
- Calculates percentage of parcel area classified as wetland or riparian zone
- Supports multiple wetland type codes from landcover rasters
- Output: 0-100% coverage
- Area-weighted calculation using coverage fractions
indicator_water_twi() (W3) - Topographic Wetness Index
- Calculates TWI using terra D8 flow algorithm
- Formula: TWI = ln(catchment_area / tan(slope))
- Automatically handles flat areas and edge cases
- Output: TWI values (typically 0-20, higher = wetter areas)
- Future: whitebox D-infinity algorithm support (v0.3.0+)

Internal Utilities

TWI Calculation System
- calculate_twi_terra() - D8 flow direction algorithm
- Slope-based flow accumulation
- Catchment area calculation
- Handles numerical edge cases (flat areas, infinite values)
- calculate_twi_whitebox() - Placeholder for future D-infinity implementation

Testing

51 new tests for water family indicators
- W1 network: 13 tests (density calculation, buffer zones, zero-stream parcels)
- W2 wetlands: 14 tests (percentage calculation, multiple codes, zero coverage)
- W3 TWI: 16 tests (DEM processing, method validation, terrain variation)
- Integration: 8 tests (combined workflow, dataframe integration)
Total test suite: 498 tests passing (up from 447)
Full backward compatibility maintained

Technical Details

W1 Network Density: Uses sf spatial operations for line-polygon intersection
W2 Wetland Coverage: Uses exactextractr for area-weighted raster value extraction
W3 TWI: Terra hydrology functions (terrain(v="flowdir"), flowAccumulation())
Flow algorithm: D8 (8-neighbor) for computational efficiency
Coordinate transformations: Automatic CRS harmonization for vector layers

v0.2.0 - Phase 4: Famille C (Carbone/Énergétique)

Status: ✅ Phase 4 Complete (447 tests passing)

New Indicator Functions

Carbon Family (Famille C) - 2 Indicators

indicator_carbon_biomass() (C1) - Aboveground carbon stock using species-specific allometric equations
- Requires: BD Forêt v2 attributes (species, age, density)
- Species support: Quercus, Fagus, Pinus, Abies, + Generic fallback
- Allometric model: Biomass = a × Age^b × Density^c
- Output: tC/ha (tonnes carbon per hectare)
- Citations: IGN/IFN literature (Dupouey, Bontemps, Vallet, Wutzler)
indicator_carbon_ndvi() (C2) - Vegetation vitality via NDVI
- Requires: Sentinel-2 or equivalent NDVI raster (0-1 scale)
- Output: Mean NDVI per forest parcel
- Future: Temporal trend analysis (v0.3.0+)

Internal Data & Utilities

Allometric Model System (R/sysdata.rda)
- 5 species-specific coefficient sets
- Calibrated for realistic French forest biomass (50-200 tC/ha mature stands)
- Source: data-raw/allometric_models.R
New Utility Functions (internal)
- get_allometric_coefficients() - Species-specific coefficient lookup
- calculate_allometric_biomass() - Vectorized biomass calculation
- detect_indicator_family() - Extract family code from indicator name
- get_family_name() - Full family name from code

Deprecations

indicator_carbon() - Now deprecated (will be removed in v1.0.0)
- Replacement: Use indicator_carbon_biomass() for BD Forêt support, or indicator_carbon_ndvi() for NDVI
- Backward compatibility: Function still works with deprecation warning
- All existing workflows continue to function

Testing

38 new tests for carbon family indicators
- C1 biomass: 15 tests (allometric calculations, NA handling, column names, Generic fallback)
- C2 NDVI: 10 tests (raster extraction, edge values, trend warning)
- Integration: 8 tests (backward compatibility, nemeton_compute integration)
- Edge cases: 5 tests (missing columns, invalid inputs, error messages)
Total test suite: 447 tests passing (up from 409)
Full backward compatibility verified

Technical Details

Allometric coefficients calibrated to produce realistic biomass values:
- Young/sparse stands: 2-10 tC/ha
- Mature forests: 50-200 tC/ha
- Age exponent (b): 1.55-1.75
- Density exponent (c): 0.80-0.90
NA propagation: Properly handles missing species, age, or density data

nemeton 0.1.0-rc1 (2026-01-04)

MVP Release Candidate

Status: ✅ 97% Complete (32/33 requirements) - Ready for testing

Major Features

Core Functionality (✅ Complete)

Spatial Analysis Engine: nemeton_compute() with 5 biophysical indicators
Automatic Preprocessing: CRS harmonization, extent cropping
Error Resilience: Per-indicator error handling (continues if one fails)
Lazy Loading: Memory-efficient layer catalog system

Indicators (✅ 5/5 Complete)

indicator_carbon() - Carbon stock from biomass (Mg C/ha)
indicator_biodiversity() - Species richness / Shannon index
indicator_water() - Water regulation (TWI + proximity to streams)
indicator_fragmentation() - Forest coverage and connectivity
indicator_accessibility() - Distance to roads and trails

Normalization & Indices (✅ Complete)

normalize_indicators() - 3 methods (min-max, z-score, quantile)
create_composite_index() - Weighted aggregation (4 methods)
invert_indicator() - Reverse polarity for negative indicators
Reference-based normalization support

Visualization (⚠️ 3/4 - Radar pending)

plot_indicators_map() - Thematic choropleth maps (single + faceted)
plot_comparison_map() - Side-by-side scenario comparison
plot_difference_map() - Absolute and relative change visualization
Multiple palettes: viridis, RdYlGn, Greens, Blues, etc.

Demo Dataset (✅ Complete)

massif_demo - Synthetic forest data (136 ha, 20 parcels)
4 rasters at 25m: biomass, DEM, landcover, species richness
2 vector layers: roads (5), water courses (3)
Lambert-93 projection (EPSG:2154)
Reproducible generation script (data-raw/massif_demo.R)

Internationalization (✅ Bonus Feature)

Bilingual Support: French + English (200+ messages)
Auto-detection: System locale detection
Manual Override: nemeton_set_language("fr") / nemeton_set_language("en")
Complete Coverage: All user-facing messages translated
Dedicated vignette: internationalization.Rmd

Exported Functions (17)

Core: nemeton_units(), nemeton_layers(), nemeton_compute(), massif_demo_layers() Indicators: indicator_carbon(), indicator_biodiversity(), indicator_water(), indicator_fragmentation(), indicator_accessibility() Normalization: normalize_indicators(), create_composite_index(), invert_indicator() Visualization: plot_indicators_map(), plot_comparison_map(), plot_difference_map() Utilities: list_indicators(), nemeton_set_language()

Documentation (✅ Complete)

README.md: Comprehensive quick start guide (497 lines)
Vignettes:
- getting-started.Rmd - Full workflow with massif_demo
- internationalization.Rmd - i18n guide (FR/EN)
Roxygen2: All 17 exported functions fully documented
Examples: Executable examples in all function docs

Testing (✅ 225+ Tests)

Unit Tests: Comprehensive coverage across all modules
Integration Tests: End-to-end workflow validation
Real Data Tests: French cadastral parcel testing
Fixtures: Helper functions for test data generation

Package Metrics

R Code: ~2,500 lines
Tests: ~2,100 lines
Dataset Size: 0.81 Mo (< 5 Mo target)
Functions: 17 exported
Vignettes: 2
i18n Messages: 200+ (FR/EN)

Quick Start Example

library(nemeton)

# 5-line workflow
data(massif_demo_units)
layers <- massif_demo_layers()
results <- nemeton_compute(massif_demo_units, layers, indicators = "all")
normalized <- normalize_indicators(results, method = "minmax")
plot_indicators_map(normalized, palette = "viridis")

Known Issues

⚠️ Minor test fixture compatibility issue (to be fixed in v0.1.0 final)
⚠️ Test coverage measurement pending (covr fails due to test issues)
📝 User Story 4 (radar chart) not implemented (P3 - optional for MVP)

Roadmap to v0.1.0

Fix test fixtures
Verify devtools::check() passes
Measure test coverage (target: ≥70%)
Optional: Implement nemeton_radar() (P3)

Breaking Changes

None (initial release)

Credits

Developed with ❤️ and Claude Code Contributors: Pascal Obstétar, Claude Sonnet 4.5

nemeton 0.14.1

UI Improvements

Bug Fixes

Documentation

nemeton 0.14.0

Test Suite Stabilization

Bug Fixes

Documentation

Tests

nemeton 0.13.0

CI/CD Optimization

nemeton 0.12.0

Phase 9 Finalization - MVP 0.7.0 Complete

New Features

Documentation

Bug Fixes

Tests

nemeton 0.8.0

New Features

nemetonApp - Interactive Shiny Application

Bug Fixes

Dependencies

nemeton 0.6.2

Changes

nemeton 0.6.1

Changes

nemeton 0.6.0 (Development)

v0.6.0 - Legacy Indicators Removal

BREAKING CHANGES ⚠️

Removed Functions

Migration Guide

Updated Core Functions

Files Removed

Rationale

nemeton 0.5.2

v0.5.2 - Tutorial 09 Sampling + TSP

New Features

Tutorial 09: Échantillonnage de calibration LiDAR HD + TSP (180 min)

Improvements

Dependencies

Documentation

nemeton 0.5.1

v0.5.1 - Tutorial 08 Coregistration

New Features

Tutorial 08: Coregistration LiDAR/Terrain (130 min)

Documentation

nemeton 0.5.0

v0.5.0 - Tutorial 07 & CRAN Compliance

Overview

New Features

Tutorial 07: LiDAR Avancé (90 min)

Dependencies

Documentation

CRAN Compliance

nemeton 0.4.0

v0.4.0 - Complete 12-Family Ecosystem Services Referential

Overview

New Indicator Families

Social & Recreational Family (Famille S) - 3 Indicators

Productive & Economic Family (Famille P) - 3 Indicators

Energy & Climate Family (Famille E) - 2 Indicators

Naturalness & Wilderness Family (Famille N) - 3 Indicators

New Analysis Functions

Multi-Criteria Decision Support

Enhanced Features

New Demo Dataset

New Vignettes

Dependencies

Documentation

nemeton 0.3.0 (Development)

v0.3.0 MVP - Multi-Family Extension (B, R, T, A)

Overview

New Indicator Families

Biodiversity Family (Famille B) - 3 Indicators

Risk & Resilience Family (Famille R) - 3 Indicators

Temporal Dynamics Family (Famille T) - 2 Indicators

Air Quality & Microclimate Family (Famille A) - 2 Indicators

Extended Functions

Testing