Tools for Tidy Functional Data • tidyfun

The goal of tidyfun is to provide accessible and well-documented software that makes functional data analysis in R easy – specifically data wrangling and exploratory analysis. This is achieved by introducing a new data type (tf). Vectors of class tf can be operated on using many standard functions (+, mean, etc.) as well as several new functions in tidyfun (tf_smooth, tf_where).

Crucially, vectors of class tf can be included in data frames containing other variables, enabling data manipulation using tidyverse tools. This approach is connected to the conceptual framework in functional data analysis, which assumes that complete functions are the unit of observation; with tidyfun, full curves sit alongside numeric, factor, and other observations on the same subject.

Installation

tidyfun is currently not on CRAN. You can install the development version from GitHub with:

# install.packages("pak")
pak::pak("tidyfun/tidyfun")

Overview

tidyfun provides:

new data types for representing functional data: tfd & tfb
arithmetic operators and descriptive statistics for such data
tidyverse-verbs for handling functional data – especially inside data frames
new graphics functions for tf vectors and tidy functional data frames

For detailed information on the features of tidyfun, check out articles on the following topics:

Representing functional data as tf vectors, and operating on those vectors
Converting non-tidy functional data (matrices, “long” and “wide” data frames) to tidy functional data
Wrangling data frames that include functional data using tidyverse and tidyfun tools
Visualizing tidy functional data

The result is a package that enables exploratory data analysis like the following, which computes group-specific mean curves in the dti_df dataset, and plots the result:

library("tidyfun")
data(dti_df, package = "tidyfun")

dti_df |>
  group_by(case, sex) |>
  summarize(mean_cca = mean(cca, na.rm = TRUE)) |>
  ggplot(aes(y = mean_cca, color = case)) +
  geom_spaghetti(size = 2) +
  facet_grid(~sex)

What does it do?

New vector-like data types for functional data

tidyfun introduces new S3-classes for functional data, either as raw data (class tfd for tidy functional data) or in basis representation (class tfb for tidy functional basis data), defined in the underlying {tf} package.

Such tf-objects can be subsetted or subassigned, computed on and summarized.

Almost all

operators like ==, + or *
math functions like sum, log or abs
and statistics functions like mean or sd

are defined for tidyfun’s data structures (more).

The tf objects are basically glorified lists, so they work well as columns in data frames. That makes it a lot easier to keep your other data and functional measurements together in one object for preprocessing, exploratory analysis and description. At the same time, these objects actually behave like vectors of functions to some extent, i.e., they can be evaluated on any point in their domain, they can be integrated or differentiated, etc.

See here for more information on the operations defined for tf vectors.

Methods for converting existing data to `tf`

tidyfun includes functions tfd and tfb for converting matrices, data frames, etc. to tf vectors. It also provides tf_gather & tf_nest in order to reshape tables with functional data, by going from wide to narrow or from long to short; functions like as.matrix, tf_spread & tf_unnest can reverse these data conversions.

See here details on getting data into (and out of) the tf format.

`tidyverse` verbs for dealing with functional data inside data frames

All dplyr verbs work on tf-columns, so you can filter, mutate, group_by & summarize, etc., functional data pretty much like conventional “tidy” data. tidyfun adds several functions that are useful in conjunction with these, like tf_anywhere and tf_smooth.

See here to see how you can wrangle functional data.

New `ggplot2` `geoms` and `stats` for functional data

tidyfun defines pasta-themed geoms for functional data:

geom_spaghetti for lines,
geom_meatballs for (lines &) points,
gglasagna for lasagna plots, with an order-aesthetic to sort the lasagna layers,
geom_capellini for glyphs plots (i.e., sparklines),

as well as new methods for base R graphics functions plot, lines and points for quick and easy visualizations of functional data.

See here for the documentation of the visualization approaches or take a look at the Visualization vignette.

Found a bug? Got a question? Missing some functionality?
Please let us know so we can make it better.

tidyfun