self-adjoin {ggalluvial}

Adjoin a dataset to itself

Description

This function binds a dataset to itself along adjacent pairs of a key variable. It is invoked by geom_flow() to convert data in lodes form to something similar to alluvia form.

Usage

self_adjoin(
  data,
  key,
  by = NULL,
  link = NULL,
  keep.x = NULL,
  keep.y = NULL,
  suffix = c(".x", ".y")
)

Arguments

data

A data frame in lodes form (repeated measures data; see alluvial-data).

key

Column of data indicating sequential collection; handled as in tidyr::spread().

by

Character vector of variables to self-adjoin by; passed to dplyr::mutate-joins functions.

link

Character vector of variables to adjoin. Will be replaced by pairs of variables suffixed by suffix.

keep.x, keep.y

Character vector of variables to associate with the first (respectively, second) copy of data after adjoining. These variables can overlap with each other but cannot overlap with by or link.

suffix

Suffixes to add to the adjoined link variables; passed to dplyr::mutate-joins functions.

Details

self_adjoin invokes dplyr::mutate-joins functions in order to convert a dataset with measures along a discrete key variable into a dataset consisting of column bindings of these measures (by any by variables) along adjacent values of key.

See Also

Other alluvial data manipulation: alluvial-data

Examples

# self-adjoin `majors` data
data(majors)
major_changes <- self_adjoin(majors, key = semester,
                             by = "student", link = c("semester", "curriculum"))
major_changes$change <- major_changes$curriculum.x == major_changes$curriculum.y
head(major_changes)

# self-adjoin `vaccinations` data
data(vaccinations)
vaccination_steps <- self_adjoin(vaccinations, key = survey, by = "subject",
                                 link = c("survey", "response"),
                                 keep.x = c("freq"))
head(vaccination_steps)
vaccination_steps <- self_adjoin(vaccinations, key = survey, by = "subject",
                                 link = c("survey", "response"),
                                 keep.x = c("freq"),
                                 keep.y = c("start_date", "end_date"))
head(vaccination_steps)