gf_linerange()
#> gf_linerange() uses 
#>     * a formula with shape ymin + ymax ~ x or y ~ xmin + xmax. 
#>     * geom:  linerange 
#>     * key attributes:  alpha, color, group, linetype, linewidth
#> For more information, try ?gf_linerange
gf_ribbon(low_temp + high_temp ~ date,
  data = mosaicData::Weather,
  fill = ~city, alpha = 0.4
  gf_theme(theme = theme_minimal())
gf_linerange(
  low_temp + high_temp ~ date | city ~ .,
  data = mosaicData::Weather,
  color = ~ ((low_temp + high_temp) / 2)
  gf_refine(scale_colour_gradientn(colors = rev(rainbow(5)))) |>
  gf_labs(color = "mid-temp")
gf_ribbon(low_temp + high_temp ~ date | city ~ ., data = mosaicData::Weather)
# Chaining in the data
mosaicData::Weather |>
  gf_ribbon(low_temp + high_temp ~ date, alpha = 0.4) |>
  gf_facet_grid(city ~ .)
if (require(mosaicData) && require(dplyr)) {
  HELP2 <- HELPrct |>
    group_by(substance, sex) |>
    summarise(
      age = NA,
      mean.age = mean(age),
      median.age = median(age),
      max.age = max(age),
      min.age = min(age),
      sd.age = sd(age),
      lo = mean.age - sd.age,
      hi = mean.age + sd.age
  gf_jitter(age ~ substance, data = HELPrct,
      alpha = 0.5, width = 0.2, height = 0, color = "skyblue") |>
    gf_pointrange(mean.age + lo + hi ~ substance, data = HELP2) |>
    gf_facet_grid(~sex)
  gf_jitter(age ~ substance, data = HELPrct,
    alpha = 0.5, width = 0.2, height = 0, color = "skyblue") |>
    gf_errorbar(lo + hi ~ substance, data = HELP2, inherit = FALSE) |>
    gf_facet_grid(~sex)
  # width is defined differently for gf_boxplot() and gf_jitter()
  #   * for gf_boxplot() it is the full width of the box.
  #   * for gf_jitter() it is half that -- the maximum amount added or subtracted.
  gf_boxplot(age ~ substance, data = HELPrct, width = 0.4) |>
    gf_jitter(width = 0.4, height = 0, color = "skyblue", alpha = 0.5)
  gf_boxplot(age ~ substance, data = HELPrct, width = 0.4) |>
    gf_jitter(width = 0.2, height = 0, color = "skyblue", alpha = 0.5)
#> `summarise()` has grouped output by 'substance'. You can override using the
#> `.groups` argument.
p <- gf_jitter(mpg ~ cyl, data = mtcars, height = 0, width = 0.15); p
p |> gf_summary(fun.data = "mean_cl_boot", color = "red", size = 2, linewidth = 1.3)
# You can supply individual functions to summarise the value at
# each x:
p |> gf_summary(fun.y = "median", color = "red", size = 3, geom = "point")
#> No summary function supplied, defaulting to `mean_se()`
  gf_summary(fun.y = "mean", color = "red", size = 3, geom = "point") |>
  gf_summary(fun.y = mean, geom = "line")
#> No summary function supplied, defaulting to `mean_se()`
#> No summary function supplied, defaulting to `mean_se()`
  gf_summary(fun.y = mean, fun.ymin = min, fun.ymax = max, color = "red")
#> No summary function supplied, defaulting to `mean_se()`
if (FALSE) {
  gf_summary(fun.ymin = min, fun.ymax = max, color = "red", geom = "linerange")
gf_bar(~ cut, data = diamonds)
gf_col(price ~ cut, data = diamonds, stat = "summary_bin", fun.y = "mean")
#> No summary function supplied, defaulting to `mean_se()`
#> No summary function supplied, defaulting to `mean_se()`
#> No summary function supplied, defaulting to `mean_se()`
#> No summary function supplied, defaulting to `mean_se()`
#> No summary function supplied, defaulting to `mean_se()`
# Don't use gf_lims() to zoom into a summary plot - this throws the
# data away
p <- gf_summary(mpg ~ cyl, data = mtcars, fun.y = "mean", geom = "point")
#> No summary function supplied, defaulting to `mean_se()`
p |> gf_lims(y = c(15, 30))
#> Warning: Removed 9 rows containing non-finite values (`stat_summary()`).
#> No summary function supplied, defaulting to `mean_se()`
# Instead use coord_cartesian()
p |> gf_refine(coord_cartesian(ylim = c(15, 30)))
#> No summary function supplied, defaulting to `mean_se()`
# A set of useful summary functions is provided from the Hmisc package.
if (FALSE) {
p <- gf_jitter(mpg ~ cyl, data = mtcars, width = 0.15, height = 0); p
p |> gf_summary(fun.data = mean_cl_boot, color = "red")
p |> gf_summary(fun.data = mean_cl_boot, color = "red", geom = "crossbar")
p |> gf_summary(fun.data = mean_sdl, group = ~ cyl, color = "red",
                   geom = "crossbar", width = 0.3)
p |> gf_summary(group = ~ cyl, color = "red", geom = "crossbar", width = 0.3,
        fun.data = mean_sdl, fun.args = list(mult = 1))
p |> gf_summary(fun.data = median_hilow, group = ~ cyl, color = "red",
        geom = "crossbar", width = 0.3)
# An example with highly skewed distributions:
if (require("ggplot2movies")) {
  set.seed(596)
  Mov <- movies[sample(nrow(movies), 1000), ]
  m2 <- gf_jitter(votes ~ factor(round(rating)), data = Mov, width = 0.15, height = 0, alpha = 0.3)
  m2 <- m2 |>
    gf_summary(fun.data = "mean_cl_boot", geom = "crossbar",
               colour = "red", width = 0.3) |>
    gf_labs(x = "rating")
  # Notice how the overplotting skews off visual perception of the mean
  # supplementing the raw data with summary statistics is _very_ important
  # Next, we'll look at votes on a log scale.
  # Transforming the scale means the data are transformed
  # first, after which statistics are computed:
  m2 |> gf_refine(scale_y_log10())
  # Transforming the coordinate system occurs after the
  # statistic has been computed. This means we're calculating the summary on the raw data
  # and stretching the geoms onto the log scale.  Compare the widths of the
  # standard errors.
  m2 |> gf_refine(coord_trans(y="log10"))
#> Loading required package: ggplot2movies