tidyfun 0.2.0

• ggplot2 support for multivariate functional data (tf_mv columns, see tf::tfd_mv()): map them with aes(tf = ...) in tf_ggplot(), with type = "facet" (value-vs-arg curves, one panel per output dimension) or type = "trajectory" (planar curves x(t) vs y(t) for 2-component objects).
• New autoplot.tf_mv() and autolayer.tf_mv() methods for quick plots and layers of tf_mv objects.
• New tf_unnest.tf_mv() method returning a “wide” long table with one value column per output dimension: (id, arg, <component 1>, ..., <component d>).
• Behavior change: an explicitly mapped x aesthetic is no longer silently overwritten by tf_y’s arg grid, so aes(tf_x = fx, tf_y = fy) now correctly draws planar curves x(t) vs y(t).
• Univariate-only displays (geom_spaghetti()/geom_meatballs(), gglasagna(), geom_fboxplot()) now fail early with an informative error for tf_mv inputs instead of dying with obscure internal errors.

tidyfun 0.1.2

• Rebuilt chf_df with current tf constructors before saving so the packaged dataset no longer carries stale namespace references from pre-tf releases.

tidyfun 0.1.1

• Initial CRAN submission.
• Functional data types (tfd, tfb) as data frame columns via the tf package.
• tidyverse-compatible data wrangling: tf_gather, tf_spread, tf_nest, tf_unnest.
• ggplot2 geoms for functional data: geom_spaghetti, geom_meatballs, geom_capellini, geom_errorband, gglasagna.
• tf_ggplot() for tf-aware ggplot construction with standard geoms.
• Functional boxplots via geom_fboxplot.