Skip to contents

Changelog

Source: NEWS.md

ggconsulting 0.1.1

This patch release follows the bundled reference corpus into theme_finance(), teaches ct_finish() to mirror axes and defer to caller-supplied scales, records dataset provenance, and starts organised documentation: a gallery article, provenance notes with source credits, and an R-universe distribution.

Data-aware polish

  • ct_finish(mirror_y = TRUE) repeats the y-axis ticks on the opposite edge of the panel, without labels, so the eye can track a level across a wide panel in the absence of gridlines. It pairs with the bare axis of theme_finance().
  • Auto expansion in ct_finish() now defers to any positional scale supplied by the caller, so scale_y_continuous(labels = ...) keeps its labels instead of being silently replaced. Set the expansion in the same call when you need both labels and headroom.

Themes

  • theme_finance() follows the institutional finance chart packs the archetype is drawn from: humanist sans typography (Source Sans 3 with an Inter / Helvetica Neue / Arial fallback chain) instead of serif, no gridlines, and y-axis ticks to read levels against. Pair with ct_finish(mirror_y = TRUE) to repeat the ticks on the right edge of wide panels.

Fonts

  • install_consulting_fonts() now installs to the per-user Windows Fonts directory by default and registers the installed fonts for the active R session.

Distribution and documentation

  • The package is available from R-universe, with source and binary builds for supported platforms.
  • The README is rewritten against the current API, with a quick start that demos ct_finish() end labels and a Why ggconsulting? summary. The Codecov badge is replaced by the R-universe version badge.
  • New article “Themes and palettes”: a gallery with one plot per archetype preset plus the discrete and continuous scales, indexed under a pkgdown Articles navbar entry. The site gains a Datasets reference section.
  • Dataset provenance is recorded in data-raw/README.md and mirrored into the dataset help pages. Beyond new disclaimers, this corrects the br_macro description (BCB SGS series 4189 and 3696; sale-rate USD/BRL; monthly source series rather than aggregated dailies) and credits B3 for ibov_sectors. Bundled snapshots note that BCB, IBGE, and B3 do not endorse the package.
  • The package has a hex logo, regenerable with data-raw/create_hexlogo.R, shown on the package help page.

Development

  • data-raw/references/ catalogues 22 institutional finance references, with McKinsey- and IMF-style briefs to guide future theme work.
  • Local agent and editor files (.Rprofile, CLAUDE.md, AGENTS.md, Playwright MCP session output, README.html) are no longer tracked.

ggconsulting 0.1.0

First public release. ggconsulting is an opinionated ggplot2 extension for executive-grade consulting output: archetype themes, palettes, scales, locale-aware label helpers, and a data-aware polish layer.

The package is experimental. The public API is being shaped against real consulting decks, so expect breaking changes through the 0.x series.

Requirements

  • Requires ggplot2 >= 4.0.0 and R >= 4.1. The themes are built on ggplot2 4.x API — theme_sub_*() helpers, element_geom(), and from_theme() linkage — and will not work on ggplot2 3.x.

Themes

  • ct_theme() composes a theme from palette, font, font_fallback, density, and context arguments. The palette’s main colour is routed through element_geom(ink = ...), so from_theme()-aware geoms pick it up without an explicit scale_color_*() call.
  • theme_strategy() — minimal, generous whitespace, navy default.
  • theme_finance() — serif preset (Source Serif 4 → Georgia → Times New Roman → serif) with a regular-weight title, lighter major gridlines, and denser defaults (density = "tight", context = "report") tuned for printed pages rather than slides. Defaults to finance_classic.
  • theme_editorial() — serif preset with an italic subtitle and a larger, tighter-leaded title. Defaults to editorial_warm.
  • ct_theme(paper = ...) sets the figure ground. Pass "cream" (Financial Times), "warm_grey" (The Economist), "white", or any colour R recognises. Because theme_minimal() leaves the panel background blank, filling the plot background alone gives a uniform ground with no panel-versus-plot seam. Setting paper also warms the major gridline to match. The archetypes forward it, so theme_editorial(paper = "cream") works; cream is opt-in rather than the editorial default.

Palettes and scales

  • Eleven palettes of six colours each, in three families: five strategy (strategy_navy, strategy_emerald, strategy_crimson, strategy_azure, strategy_slate), three finance (finance_classic, finance_steel, finance_burgundy), and three editorial (editorial_warm, editorial_clay, editorial_oxide).
  • ct_palette() returns a palette’s colours, subsets to n, or — called with no arguments — lists every available palette name.
  • scale_color_ct() / scale_fill_ct() — discrete scales. When the data needs more levels than the palette holds, colours are interpolated and a warning points at the continuous scales instead.
  • scale_color_ct_c() / scale_fill_ct_c() — continuous variants, with direction = -1 to reverse.
  • British-spelling aliases: scale_colour_ct(), scale_colour_ct_c().
  • ct_palette_show() — swatch preview for one palette, a custom hex vector, or every shipped palette faceted.

Locale and formatters

  • ct_locale("pt-BR" | "en-US") — session-scoped locale switch stored in options(ggconsulting.locale). It does not touch Sys.setlocale(); the package ships its own month tables and formatting marks, so output is identical across Windows, Linux, and macOS.
  • fmt_number() — locale-aware numbers (1.234,5 / 1,234.5).
  • fmt_brl() — Brazilian Real, always R$ with a non-breaking space regardless of the active locale. style = "accounting" wraps negatives in parentheses.
  • fmt_currency() — uses the active locale’s currency symbol.
  • fmt_pct() — fraction to percent (0.5 → "50%").
  • fmt_delta() — always-signed percentage-point deltas (+1,2pp).
  • fmt_month() — Date / POSIXct to a localised month string.

Data-aware polish

  • ct_finish() runs after the geom layer is built, via a ggplot_add() S3 method, and inspects the plot to apply:
    • value labels above bars or beside points (values = TRUE)
    • reordering of a categorical x by y (sort = "asc" | "desc")
    • label formatting by shortcut (label_fmt = "brl" | "number" | "pct" | "delta") or a user-supplied function
    • highlighting of specific x values in the theme’s main colour, muting the rest with muted_color (default #A8A4A0, a warm-leaning neutral that reads against both cool and warm palettes)
    • end labels on the last point of each line series (end_labels = TRUE), or only in the first panel of a faceted plot (end_labels = "first_facet"), which replaces a legend across small multiples
    • a filled point at each series’ last observation (end_points = TRUE)
    • a right-hand y axis (axis_y = "right"), applied through guide_axis() so it leaves a scale_y_*() call of your own alone
    • geom-aware scale expansion (expand = "auto") — headroom above columns, right-side room for line endpoints
    end_labels now accepts Date and POSIXct x aesthetics, which it previously rejected. Auto expansion no longer adds a second x scale over the one end labels set, and neither replaces an x scale supplied by the caller.

Geom wrappers and defaults

Fonts

  • install_consulting_fonts() downloads the five archetype font families from Google Fonts (Inter, Source Sans 3, Lato, Source Serif 4, IBM Plex Sans — all OFL or Apache-2.0) and installs them into a platform-appropriate user font directory.

    Because that writes outside the R session, it asks for confirmation first when run interactively and refuses to run otherwise. Pass an explicit dest, or set options(ggconsulting.font_consent = TRUE), to install unattended.

  • has_font() reports whether a family is available, checking both fonts installed on the operating system and fonts registered for the session with systemfonts::register_font(). This matters for client brand fonts, which are commonly registered from a file rather than installed: ct_theme(font = "ClientSans") now honours a registered ClientSans instead of falling through to the font_fallback chain.

Data

  • Six bundled datasets for examples and tests: bu_quarterly, market_share, client_nps, ebitda_bridge, ibov_sectors, and br_macro.

Development

  • inst/gallery/ ships example scripts for visual QA of themes, palettes, geoms, and formatters. Available post-install via system.file("gallery", package = "ggconsulting").