wcswatin 0.2.0

Breaking changes

• Removed the superseded raster package dependency and standardized spatial raster workflows on terra and SpatRaster.
• ts_to_area() now returns a multi-layer SpatRaster and preserves input file stems as layer names, such as 2020-01-01.
• input_raster() no longer accepts legacy RasterLayer, RasterStack, or RasterBrick objects. Convert these objects with terra::rast() before passing them to wcswatin.

wcswatin 0.1.1

Maintenance

• Minor maintenance release following CRAN pre-submission checks.
• Added package-specific terms to inst/WORDLIST.
• Excluded cran-comments.md from the source package.
• Updated package metadata and README citation examples to version 0.1.1.

wcswatin 0.1.0

Package workflows and structure

• Reorganized R source files by logical domain, making the package structure match the main workflows more closely.
• Reorganized the test suite by the same logical domains and added shared test helpers for temporary files and directories.
• Added compact NetCDF example files to support runnable examples, metadata inspection, and raster workflow tests.
• Replaced the previous Makefile-based workflow with a justfile for common development tasks.

New features

• Added raster_info() to summarize raster and NetCDF metadata, including variables, units, layer counts, spatial extent, CRS, and time range.
• Added ts_point_to_files() to save ts_to_point() outputs as individual SWAT-style point files.
• Added value_at_hour support to daily_aggregation() for products whose daily value is stored at a specific hour.
• Added value_at_hour, date_shift_days, and drop_first_layer support to datacube_aggregation() so hourly raster cubes can be reduced before extraction.
• Improved tbl_from_references() so reference points can be supplied using the package’s standard table, vector, and path input patterns.

Breaking changes

• Standardized longitude column names to LON.
• Renamed the take_out_first_record argument in daily_aggregation() to drop_first_record.
• Renamed the negatif_number argument in station/table preprocessing to neg_to_zero.

Bug fixes and validation

• Added reusable input validation helpers and clearer error messages across raster, table, station, and interpolation workflows.
• Ensured variable derivation helpers create output directories consistently.
• Removed a duplicate internal file_name() implementation.
• Removed the vroom dependency from the main file-reading workflow.
• Expanded test coverage for raster aggregation, cube extraction, metadata, reference extraction, station input, table aggregation, table I/O, variable derivation, and trend-surface interpolation.

Documentation

• Added a lightweight executable vignette using bundled example data.
• Added precomputed pkgdown articles for:
  • ERA5-Land hourly data to SWAT inputs;
  • station interpolation workflow;
  • running a similar external case study.
• Updated README and pkgdown navigation to reflect the current package workflows.

wcswatin 0.0.1

First Release

Initial release of wcswatin (Weather & Climate SWAT INput) package for preparing weather and climate data for input in the Soil & Water Assessment Tool (SWAT).

Features

• Process NetCDF and GeoTIFF raster files from multiple climate data providers
• Temporal and spatial aggregation tools
• Gap filling and missing data handling
• Unit conversion utilities for weather variables
• Summary statistics and visualization functions
• File management and batch processing utilities

Infrastructure

• Comprehensive test suite with cross-platform support (Windows, macOS, Linux)
• CI/CD with GitHub Actions (R-CMD-check, test-coverage, pkgdown)
• Documentation website available at https://reginalexavier.github.io/wcswatin/