This function allows the user to choose from a menu of prepackaged plotting styles (or to pass your own) that are designed to facilitate data comparison between groups. Plot functions are specific to one type of data (numeric or categorical) and will only use those columns when building plots. For example, histograms are only useful for numeric data, so categorical type data is ignored.
build_plots(
stack,
plot_fun,
...,
group_var = "flyover_id_",
keep_type = NULL,
ncores = 1,
plot_mods = NULL
)Tabular data that inherits from data.frame.
This is recommended to be the output of
stack_data but can be built by the user.
It is coerced to a tibble.
The flyover plotting function to apply to the
data. The user may also supply a custom function,
but must be careful to match that function with
the appropriate keep_type argument.
Note that this function must be passed without parentheses.
Additional arguments to pass to the geom of the
supplied flyover plot function. Use this for further
modifications to the plots if needed.
Character string; the column name that represents the
source of the data. It will be used as a grouping
variable in the subsequent plots. The default value
comes from the grouping column created by default in
stack_data.
Depending on the type of plot desired,
only numeric or categorical data can be used.
By default the column type is determined by the
flyover plotting function passed to the
plot_fun argument. If a flyover plot
is passed, this argument is ignored. However, if the user
specifies a custom plotting function, the
keep_type argument must be set to one of
"numeric", "categorical", or "both".
Passing "categorical" will keep character, factor,
and logical column types.
Number of cores to use for processing plots. Plot creation can
be parallelized by setting ncores > 1. Note this uses
the built-in parallel package which does not support
Windows, so this argument is ignored for Windows systems.
List containing additional layers to the ggplot2 call
such as theme changes, different color scales, etc.
Each layer should be a separate list element.
See ?ggplot2::`+.gg` for more details.
A tibble containing, for each relevant variable of the
input data, a row with a plot object and data frame of
cognostics for the trelliscope display. The tibble can
then be passed to build_display to create the
trelliscope output.