There are two primary functions in ggvis that are used to create plots: qvis and ggvis. qvis provides a streamlined interface which is suitable for simple plots; when you need more control over plots, ggvis may be more appropriate.
In some of the simpler examples below, we’ll show equivalent code qvis and ggvis.
First, load the ggvis package:
library(ggvis)We’ll use the built-in mtcars data set, and look at two columns of interest, mpg, and wt:
# The first few rows of mtcars
head(mtcars)#> mpg cyl disp hp drat wt qsec vs am gear carb
#> Mazda RX4 21.0 6 160 110 3.90 2.620 16.46 0 1 4 4
#> Mazda RX4 Wag 21.0 6 160 110 3.90 2.875 17.02 0 1 4 4
#> Datsun 710 22.8 4 108 93 3.85 2.320 18.61 1 1 4 1
#> Hornet 4 Drive 21.4 6 258 110 3.08 3.215 19.44 1 0 3 1
#> Hornet Sportabout 18.7 8 360 175 3.15 3.440 17.02 0 0 3 2
#> Valiant 18.1 6 225 105 2.76 3.460 20.22 1 0 3 1A basic scatter plot:
# qvis(mtcars, ~wt, ~mpg)
ggvis(mtcars, props(x = ~wt, y = ~mpg)) + mark_point()Smaller points, a different shape, a different outline (stroke) color, and empty fill:
# qvis(mtcars, ~wt, ~mpg, size := 25, shape := "diamond", stroke := "red", fill := NA)
ggvis(mtcars, props(x = ~wt, y = ~mpg)) +
mark_point(props(size := 25, shape := "diamond", stroke := "red", fill := NA))We’ll use the built-in pressure data set for these examples:
# The first few rows
head(pressure)#> temperature pressure
#> 1 0 0.0002
#> 2 20 0.0012
#> 3 40 0.0060
#> 4 60 0.0300
#> 5 80 0.0900
#> 6 100 0.2700When the variable on the x axis is continuous (e.g., numeric or date-time):
# Note that y2 is the bottom of the bar, and the width is specified in pixels
ggvis(pressure, props(x = ~temperature, y = ~pressure, y2 = 0)) +
mark_rect(props(width := 10))In the previous example, the width of each bar was specified in pixels. If you change the width of the plot, you’ll see that the bars remain 10 pixels wide.
Instead of setting the width in pixels, it’s possible to set the width in the scaled data space, by specifying the x and x2 values for each bar. If you do this, the width of the bars isn’t determined by the In this example, each bar is 20 wide in the x scale, so the bars touch:
ggvis(pressure,
props(x = ~temperature - 10, x2 = ~temperature + 10, y = ~pressure, y2 = 0)) +
mark_rect() +
guide_axis("x", title = "temperature")When the variable on the x axis is categorical (e.g., factor or character):
# First, modify the pressure data set so that the x variable is a factor
pressure2 <- pressure
pressure2$temperature <- factor(pressure2$temperature)
# qvis(pressure2, ~temperature, ~pressure, y2 = 0, width = band(), layers = "rect") +
# dscale("x", "nominal", padding = 0, points = FALSE)
ggvis(pressure2, props(x = ~temperature, y = ~pressure)) +
mark_rect(props(y2 = 0, width = band())) +
dscale("x", "nominal", padding = 0, points = FALSE)(Notice that in this example, the x values aren’t sorted quite right: they’re sorted lexically, by the first, second, and then third digit, instead of by the numeric value. This is due to a bug in ggvis.)
# qvis(pressure, ~temperature, ~pressure, layers = "line")
ggvis(pressure, props(x = ~temperature, y = ~pressure)) + layer_line()Lines with points:
# qvis(pressure, ~temperature, ~pressure, layers = c("line", "point"))
ggvis(pressure, props(x = ~temperature, y = ~pressure)) +
layer_line() +
mark_point()We’ll use the built-in faithful data set for these examples:
# The first few rows
head(faithful)#> eruptions waiting
#> 1 3.600 79
#> 2 1.800 54
#> 3 3.333 74
#> 4 2.283 62
#> 5 4.533 85
#> 6 2.883 55Basic histogram:
# qvis(faithful, ~eruptions)
ggvis(faithful, props(x = ~eruptions)) + layer_histogram()Modify the fill color and binwidth, and add titles for the axes, since the automatic titles aren’t very informative:
ggvis(faithful, props(x = ~eruptions, fill := "#fff8dc")) +
layer_histogram(binwidth = 0.25) +
guide_axis("x", title = "eruptions") +
guide_axis("y", title = "count")Adding a smoothing line (defaults to a loess model):
# qvis(mtcars, ~wt, ~mpg, layers = c("point", "smooth"))
ggvis(mtcars, props(x = ~wt, y = ~mpg)) +
mark_point() +
layer_smooth()With a linear model, and 95% confidence interval for the model:
# qvis(mtcars, ~wt, ~mpg) + layer_smooth(method = "lm", se = TRUE)
ggvis(mtcars, props(x = ~wt, y = ~mpg)) +
layer_point() +
layer_smooth(method = "lm", se = TRUE)Coloring points by a variable:
ggvis(mtcars, props(x = ~wt, y = ~mpg)) +
layer_point(props(fill = ~factor(cyl)))Coloring points, and adding a smoother for each group. The grouping variable (which is applied before the transform_smooth is calculated) must be specified with by_group():
ggvis(mtcars, by_group(cyl), props(x = ~wt, y = ~mpg)) +
layer_point(props(fill = ~factor(cyl))) +
layer_smooth(method = "lm", props(stroke = ~factor(cyl)))The following plots won’t display correctly in this online document, but they will display correctly when run from R.
Scatter plot with a loess smoother, and span controlled by a slider:
ggvis(mtcars, props(x = ~wt, y = ~mpg)) +
layer_smooth(se = TRUE,
span = input_slider(min = 0.3, max = 1, value = 0.8, step = 0.1,
label = "Smoothing span")) +
layer_point() +
guide_axis("x", title = "Weight") +
guide_axis("y", title = "MPG")A kernel density estimate where the adjust parameter is controlled by a slider, and kernel is controlled by a select input.
ggvis(mtcars, props(x = ~wt)) +
layer_density(
adjust = input_slider(.1, 2, value = 1, step = .1, label = "Bandwidth adjustment"),
kernel = input_select(
c("Gaussian" = "gaussian",
"Epanechnikov" = "epanechnikov",
"Rectangular" = "rectangular",
"Triangular" = "triangular",
"Biweight" = "biweight",
"Cosine" = "cosine",
"Optcosine" = "optcosine"),
label = "Kernel")
)