ncmR 0.4.0 (2026-10-22)

NEW FEATURES

Added pie_plot() to show Above, Below and Neutral taxon proportions, with external labels and native ggplot2 theme support.
Added run_app(ncm = ...) to open existing NCM or NCM_group results directly in Figure studio.
Redesigned the Shiny app with a model workspace, live figure preview, chart-specific theme editing, and font and color controls.
Added figure exports in PDF, SVG, PNG, TIFF and JPEG formats, plus reproducible R recipe downloads.

scatter_plot() now uses base_theme and native ggplot2 theme objects for styling. Standalone font, axis-size and legend-placement arguments have been removed; use theme settings instead.
The combined model in grouped results is now stored under .ncm_total instead of all. This name is reserved and appears as “Total model” in Shiny.

Fixed custom start, lower and upper argument forwarding, and improved retries and diagnostics for non-converged fits.
Corrected frequency-interval boundary handling and tolerance for numerical roundoff in pie plots.
Preserved numeric and character group labels without overwriting the combined model.
Fixed missing group labels causing unrelated, explicitly selected groups to fail. Samples with missing labels are excluded from individual group fits and retained in the combined fit.
Improved automatic figure initialization and consistency between manual previews, exported figures and R recipes.

Fixed row name loss when converting data frame columns to numeric during file upload.

scatter_plot: removed Unicode characters from custom legend labels for broader compatibility.

print.unicode_ggplot() and ggsave_unicode(): no longer needed after removing Unicode dependency from plotting functions.

Added function scatter_plot: NCM scatter plot with fitted curve and CI.
Added function print.unicode_ggplot: Enables Unicode display when printing plots.
Added function ggsave_unicode: Saves plots with full Unicode support.