Skip to contents

Several views of a survey, side by side

Source: R/plot-panel.R
plot_survey_panel.Rd

Draws more than one plot_survey() view on one figure, so the stages can be compared without holding two windows next to each other. Optionally writes one figure per survey day.

Usage

plot_survey_panel(
  dat,
  views = NULL,
  by_day = FALSE,
  dir = "survey-panels",
  ncol = NULL,
  width = 14,
  height = 9,
  dpi = 120,
  dates = NULL,
  years = NULL,
  months = NULL,
  ...
)

Arguments

dat

A survey data frame at any stage.

views

Which views to draw. Defaults to the ones the data supports, in pipeline order: a view whose column is absent is skipped rather than erroring, because the point is to see how far the data has got.

by_day

Write one figure per survey day instead of returning one figure for everything? Default FALSE.

dir

Directory the per-day figures are written to. Default "survey-panels", created if needed. It is deliberately not the working directory: by_day = TRUE over a season writes one file per survey day.

ncol

Panels per row. Defaults to a layout chosen from how many views there are, which is what you want unless a particular figure needs a shape.

width, height, dpi

Size of the written figures, in inches and dots per inch. Only used when by_day = TRUE; otherwise the figure is returned and the caller decides.

dates, years, months

Narrow which survey days are drawn, as in filter_days(). They narrow together rather than adding up: years = 2019 with months = 8 is August 2019, not all of 2019 and every August.

...

Passed to plot_survey() for every view - coastline, sightings, max_points and the rest. A coastline resolved once and passed in is much faster than coastline = TRUE, which refetches Natural Earth for every panel of every day.

Why side by side

The views answer each other. "raw" shows every position the file holds; "platform" shows which of them this package will keep; "effort" shows which of those count; "occupations" shows how they were divided into lines. A record that vanishes between two panels is the question worth asking, and it is far easier to see in one figure than in four.

One figure per day

by_day = TRUE writes a figure for each survey day into dir. A season on one map is a smear — 24,795 line occupations over 187 days tell you nothing about any of them — while a day is a handful of lines you can actually check. The files are named survey-<date>.png and the paths are returned, so a targets pipeline can track them.

See also

plot_survey(), which draws one view.