7.5 ggplot

Reference: https://ggplot2.tidyverse.org/reference/index.html

Change global settings

theme_set(theme_bw()) # change default theme to theme_bw() globally

You start with ggplot(), supply a dataset and aesthetic mapping (with aes()). You then add on layers (like geom_point() or geom_histogram()), scales (like scale_colour_brewer()), faceting specifications (like facet_wrap()) and coordinate systems (like coord_flip()).

Default theme: theme_gray() grey background and white gridlines.

ggplot(data = NULL, mapping = aes(), ...) is used to construct the initial plot object, and is almost always followed by a plus sign (+) to add components to the plot. Doc.

The data = and mapping = specifications in the arguments are optional (and are often omitted in practice), so long as the data and the mapping values are passed into the function in the right order.

There are three common patterns used to invoke ggplot():

• ggplot(data = df, mapping = aes(x, y, other aesthetics))

  Recommended if all layers use the same data and the same set of aesthetics, although this method can also be used when adding a layer using data from another data frame.

• ggplot(data = df)

  This is useful when one data frame is used predominantly for the plot, but the aesthetics vary from one layer to another. Different x/y-axis.

• ggplot()

  This is useful when multiple data frames are used to produce different layers, as is often the case in complex graphics.

Create a customized figure theme code snippet that could be used repetitively.

theme() note that when you call it, just use + mytheme w/o parentheses as it is not a function — it is a theme setting.

mytheme <- theme(legend.position = "none", # disable legend
                 legend.spacing.y = unit(0, 'mm'), # spacing between legend title and legend items
                         legend.key.height = unit(0.8,"line"), # vertical spacing between legend items
                         legend.margin = margin(t=0, b=0, unit="mm"), # legend box margins
                 title = element_text(size=8),
                 axis.title = element_text(size=rel(1.2)), # use `rel()` to change proportionally to base font size; or a number to specify absolute size as follows;
                 axis.text = element_text(size=8), # tick labels along axes
                 panel.grid.minor = element_blank() # remove minor gridlines
                )
                
# p is a ggplot() subject
p + mytheme

rel(x) specify sizes relative to the parent.

theme_bw(base_size = 14) the default font size is 11 pt, which can be too small. Set base_size=14 to enlarge the font size for a specific theme.

theme_get() returns the current active theme.

# get default values of theme parameters
theme_get()$plot.margin 

Check default options for a specific theme

theme_bw()$legend.spacing

Legend layout

p + guides(colour = guide_legend(nrow = 1))

---

Theme

Explore themes: https://ggplot2.tidyverse.org/reference/ggtheme.html

default: theme_gray() (with grey background)

normal: theme_bw() (recommended 结构清晰)

coordinates: theme_minimal() (no axis borders 极简风) 没有背景填充色，深色模式下预览困难。

with no grid: theme_classic() (hard to read)

---

7.5.1 Wide table to Long table

tidyr::gather(data, key = "key", value = "value", ...) convert data frame to key-value long format.

• ... is a selection of columns. If empty, all variables are selected. You can supply bare variable names, select all variables between x and z with x:z, exclude y with -y.

• key one identifier column that you use to identify groups, store the names of columns that you want to gather/stack;

• value one value column name that put values in key columns;

• ... specification of columns to gather/stack. Allowed values are:

  • variable names, put them just in a sequence, do not need to wrap it in a vector

  • if you want to select all variables between a and e, use a:e. Or could use position index e.g., 1:3.

  • if you want to exclude a column name y use -y, usually the index/identifier of the columns.

df %>% pivot_longer(c(x, y, z), names_to = "key", values_to = "value")

is equivalent to df %>% gather("key", "value", x, y, z), more recommended as gather is deprecated now.

---

Long table to wide table

spread, reverses gather()

spread(data, key, value, fill = NA, convert = FALSE, drop = TRUE, sep = NULL)

• key the column in the long table that will become factor/columns in wide table; the number of new columns equals to the number of categories in key;
• the columns are filled by value;
• fill what value to fill in the new columns when there is no data for that category; default to NA;

It is recommended switching to pivot_wider(), which is easier to use, more featureful, and still under active development.

pivot_wider(data, names_from = key, values_from = value, names_prefix = "")

• names_prefix optional prefix to add to the names of the new columns, useful when you want to avoid name conflicts with existing columns.

df %>% spread(key, value) 
# equivalent to 
df %>% pivot_wider(names_from = key, values_from = value)

df <- data.frame(month=rep(1:3,2),
                 student=rep(c("Amy", "Bob"), each=3),
                 A=c(9, 7, 6, 8, 6, 9),
                 B=c(6, 7, 8, 5, 6, 7))

#       month student A B
# 1     1     Amy   9 6
# 2     2     Amy   7 7
# 3     3     Amy   6 8
# 4     1     Bob   8 5
# 5     2     Bob   6 6
# 6     3     Bob   9 7

## 1. construct long table
df %>% 
  gather(variable, value, -(month:student))
#       month student variable value
# 1      1     Amy        A     9
# 2      2     Amy        A     7
# 3      3     Amy        A     6
# 4      1     Bob        A     8
# 5      2     Bob        A     6
# 6      3     Bob        A     9
# 7      1     Amy        B     6
# 8      2     Amy        B     7
# 9      3     Amy        B     8
# 10     1     Bob        B     5
# 11     2     Bob        B     6
# 12     3     Bob        B     7

## 2. create identifier column that will become columns in the `spread` step
df %>% 
  gather(variable, value, -(month:student)) %>%
  unite(temp, student, variable)
#    month  temp value
# 1      1 Amy_A     9
# 2      2 Amy_A     7
# 3      3 Amy_A     6
# 4      1 Bob_A     8
# 5      2 Bob_A     6
# 6      3 Bob_A     9
# 7      1 Amy_B     6
# 8      2 Amy_B     7
# 9      3 Amy_B     8
# 10     1 Bob_B     5
# 11     2 Bob_B     6
# 12     3 Bob_B     7

## 3. `spread` long table, spread the column containing the factor/identifier info
df %>% 
  gather(variable, value, -(month:student)) sub%>%
  unite(temp, student, variable) %>% # first unite
  spread(temp, value) # then spread
  
#   month Amy_A Amy_B Bob_A Bob_B
# 1     1     9     6     8     5
# 2     2     7     7     6     6
# 3     3     6     8     9     7

unite(data, col, ..., sep = "_", remove = TRUE, na.rm = FALSE) paste together multiple columns into a single column or variable (often used as factor identifier for long table); return a string column.

annotate(geom = 'text', label = 'Africa', x = Inf, y = Inf, hjust = 2, vjust = 2) Add text/annotation to a designated position.

• Justification (hjust, vjust): Horizontal and vertical justification have the same parameterisation, either a string (“top”, “middle”, “bottom”, “left”, “center”, “right”) or a number between 0 and 1:

  • top = 1, middle = 0.5, bottom = 0
  • left = 0, center = 0.5, right = 1
  • Note that you can use numbers outside the range (0, 1), but it’s not recommended.

• If label is a TeX expression, then should use

   annotate("text", x=3, y=40, label=TeX(eqn, output="character"),
             hjust=0, size = 4, parse = TRUE)

---

7.5.2 Dual y-axis plot

rescaleY <- function(y1, y2){
  # useful for plotting figures with dual axis but with different range
  # y1 is the primary axis    # target to
  # y2 is the secondary axis  # origin from
  # rescale y2 to match y1 range
  # return a-intercept, b-slope
  ylim1 <- c(min(y1), max(y1))  
  ylim2 <- c(min(y2), max(y2))
  b <- (ylim1[2]-ylim1[1])/(ylim2[2]-ylim2[1])
  a <- ylim1[1]-b*ylim2[1]
  return (c(a,b)) 
}

scaleFactor <- rescaleY(the_model_df$temp, the_model_df$rsds)  # rescale rsds to match the range with temp

dev.new()
p <- ggplot(the_model_df, aes(x = yr)) +
  geom_line(aes(y = temp, colour = "tmp")) + # bluish
  geom_line(aes(y = scaleFactor[1]+rsds*scaleFactor[2], colour = "rsds")) + # orange
  scale_color_discrete(name = "Y series", # legend name
                       values = c("tmp"="#00BFC4", "rsds"="#F8766D"), # named vector for color scale, specifying colors you want for each series
                       breaks = c("tmp", "rsds"), # specify order of scales showing up in legend
                       labels = c("temperature", "radiation") # names/text show up at scales
                      ) +
  scale_y_continuous(name = "Temperature [K]",
    sec.axis = sec_axis(~(.-scaleFactor[1])/scaleFactor[2], name = "Radiation [Wm^-2]")) +
  labs(x='year', title=the_model) +
  theme(legend.title = element_blank(), # remove legend title
        legend.text = element_text(size=8), # change legend font size to smaller
        # axis.title.x =  element_blank(), # remove x axis title
        axis.title.y.right=element_text(color="#F8766D"), # y axis label
        axis.text.y.right=element_text(color="#F8766D"))
p

# save to file
ggsave("name.png", path=fig_dir, width=8.93, height=5.74, units="in", dpi=300)

f_name <- paste0("./figures/", "pre_2.png")
f_name
ppi <- 300
png(f_name, width=7.96*ppi, height=4.19*ppi, res=ppi)
print (p)
dev.off()


plot_png <- function(p, f_name, width, height, ppi=300){
    # a plot wrapper
    png(f_name, width=width*ppi, height=height*ppi, res=ppi)
    print (p)
    invisible(dev.off())
}

The second axis could be either the second x or y axis.

sec.axis(trans = NULL,name = waiver(), breaks = waiver(), labels = waiver(), guide = waiver())

• trans A formula or function of transformation from right to left.
• name The name/title of the secondary axis.
• labels A character vector giving labels (must be same length as breaks); or a function that takes the breaks as input and returns labels as output

dup.axis(trans = ~.,name = waiver(), breaks = waiver(), labels = waiver(), guide = waiver())

is provide as a shorthand for creating a secondary axis that is a duplication of the primary axis, effectively mirroring the primary axis.

dual axes rescale

for dual axes, a major problem is to rescale the second axis so that two axes could be visualized normally.

https://stackoverflow.com/questions/31953747/r-ggplot2-dual-y-axis-facet-wrap-one-histogram-and-other-line

scale_*_log10(), scale_*_sqrt() and scale_*_reverse() are useful for axis transformation.

The principal components of every plot can be defined as follow:

• data is a data frame
• Aesthetics is used to indicate x and y variables. It can also be used to control the color, the size or the shape of points, the height of bars, fill (“inside” color), linetype, etc…..
• Geometry defines the type of graphics (histogram, box plot, line plot, density plot, dot plot, ….)

In most cases you start with ggplot(), supply a dataset and aesthetic mapping (with aes()). You then add on layers (like geom_point() or geom_histogram()), scales (like scale_colour_brewer()), faceting specifications (like facet_wrap()) and coordinate systems (like coord_flip()).

library(ggplot2)
ggplot(mpg, aes(displ, hwy, colour = class)) +  # plot
  geom_point() # layer

aes(x, y, ...) function that sets aesthetic mappings. aes() can be specified either in plot or in layers.

• color when in bar plot, color means for border, if you want to sepecify bar, you should use fill

7.5.3 aes()

Aesthetic mappings aes(x, y, ...) describe how variables in the data are mapped to visual properties (aesthetics) of geoms. Aesthetic mappings can be set in ggplot() and in individual layers.

aes() is a quoting function. This means that its inputs are quoted to be evaluated in the context of the data. This makes it easy to work with variables from the data frame because you can name those directly. The flip side is that you have to use quasiquotation to program with aes(). See a tidy evaluation tutorial such as the dplyr programming vignette to learn more about these techniques.

Specifying the aesthetics in the plot vs. in the layers

https://ggplot2-book.org/layers.html

If you only have one layer in the plot, the way you specify aesthetics doesn’t make any matter. However, the distinction is important when you start adding additional layers. These two plots are both valid and interesting, but focus on quite different aspects of the data:

ggplot(mpg, aes(displ, hwy, colour = class)) + 
  geom_point() + 
  geom_smooth(se = FALSE)

ggplot(mpg, aes(displ, hwy)) + 
  geom_point(aes(colour = class)) + 
  geom_smooth(se = FALSE)

Not easy way to know the most correct method, only trial and error… 😌

Generally, you want to set up the mappings to illuminate the structure underlying the graphic and minimise typing. It may take some time before the best approach is immediately obvious, so if you’ve iterated your way to a complex graphic, it may be worthwhile to rewrite it to make the structure more clear.

Default method for geom_smooth: spline smoothing.

geom_smooth(stat = 'smooth', color = 'Red', method = 'gam', formula = y ~ s(x, bs = "cs"))

Setting vs. mapping

Instead of mapping an aesthetic property to a variable, you can set it to a single value by specifying it in the layer parameters.

We

• map an aesthetic to a variable (e.g., aes(colour = cut)) or

  • mapping will generate each layer for every class in cut

• set it to a constant (e.g., colour = "red"). (See par() for a mapping list)

  • assign values for aesthetics

    # specifying aesthetics inside `aes()` or outside is very different
    
    # outside `aes()`, passing aesthetic arguments by variable-value pairs
    ggplot(mpg, aes(cty, hwy)) + 
      geom_point(colour = "darkblue") 
    
    # inside `aes()`
    # maps (not sets) the colour to the value ‘darkblue’. This effectively creates a new variable containing only the value ‘darkblue’ and then scales it with a colour scale. Because this value is discrete, the default colour scale uses evenly spaced colours on the colour wheel, and since there is only one value this colour is pinkish.
    
    ggplot(mpg, aes(cty, hwy)) + 
    ggplot(mpg, aes(cty, hwy)) + 
      geom_point(aes(colour = "darkblue"))

    

  A third approach is to map the value, but override the default scale:

  ggplot(mpg, aes(cty, hwy)) + 
    geom_point(aes(colour = "darkblue")) + 
    scale_colour_identity() # Use this set of scales when your data has already been scaled, i.e. it already represents aesthetic values that ggplot2 can handle directly. 

  The functions scale_colour_identity(), scale_fill_identity(), scale_size_identity(), etc. work on the aesthetics specified in the scale name: colour, fill, size, etc. Use this set of scales when your data has already been scaled, i.e. it already represents aesthetic values that ggplot2 can handle directly. This is most useful if you always have a column that already contains colours.

  • assign names for multiple layers

    It’s sometimes useful to map aesthetics to constants. For example, if you want to display multiple layers with varying parameters, you can “name” each layer through “color/colour”:

    ggplot(mpg, aes(displ, hwy)) + 
      geom_point() +
      geom_smooth(aes(colour = "loess-name"), method = "loess", se = FALSE) + # a layer called "loess-name", generating a colour scale named "loess-name", not actually assigning colors to the series. Need to assign color through a color scale, ex. `scale_color_discrete()`.
      geom_smooth(aes(colour = "lm-name"), method = "lm", se = FALSE) # a layer called "lm-name"

aes() also standardises aesthetic names by converting color to colour (also in substrings, e.g., point_color to point_colour) and translating old style R names to ggplot names (e.g., pch to shape and cex to size).

Continuous scales for data (customize x & y axis)

Define your own axis preferences, breaks, limits, labels, …

scale_x_continuous() and scale_y_continuous() are the default scales for continuous x and y aesthetics. There are three variants that set the trans argument for commonly used transformations: scale_*_log10(), scale_*_sqrt() and scale_*_reverse().

scale_x_continuous(name = waiver(),breaks = waiver(), minor_breaks = waiver(),n.breaks = NULL,labels = waiver(), limits = NULL,expand = waiver(),oob = censor, na.value = NA_real_, trans = "identity", guide = waiver(),position = "bottom", sec.axis = waiver() )

• limits 
  • NULL to use the default scale range.
  • e.g., c(0,1) A numeric vector of length two providing limits of the scale. Use NA to refer to the existing minimum or maximum.
  • A function that accepts the existing (automatic) limits and returns new limits. Also accepts rlang lambda function notation. Note that setting limits on positional scales will remove data outside of the limits.
    • If the purpose is to zoom, withouting clipping, use the limit argument in the coordinate system (see coord_cartesian(xlim=c(0, 100), ylim=c(10, 20))).
  • alternatively, could use + xlim(0, 100) + ylim(10, 20) to achieve the same effects. This is the same as scale_x_continuous(limits=c(0, 100)) + scale_y_continuous(limits=c(10, 20))
• breaks
  • NULL for no breaks
  • waiver() for the default breaks computed by the transformation object
  • A numeric vector of positions. Note that the vector will be cut off if the range exceeds the data coverage.
  • A function that takes the limits as input and returns breaks as output (e.g., a function returned by scales::extended_breaks()). Also accepts rlang lambda function notation.

Other position scales: scale_x_binned(), scale_x_date(), scale_x_discrete().

scale_x_date() : class Date

scale_x_datetime() : class POSIXct

scale_x_datetime(labels = scales::date_format("%Y", tz = "CET"),
                 breaks = seq(as.POSIXct("1960-12-31 01:00:00 CET"),
                              as.POSIXct("2015-02-11 01:00:00 CET"), "10 years")
                 )

• scales::date_format("%Y", tz = "CET") is a wrapper for formatting dates on the axis.

scale_x_time() : class hms

scale_(x|y)_binned() are scales that discretize continuous position data. You can use these scales to transform continuous inputs before using it with a geom that requires discrete positions. An example is using scale_x_binned() with geom_bar() to create a histogram.

scale_*_manual(..., values, breaks = waiver()) specify your own set of mappings from levels in the data to aesthetic values.

• * could be one of color, fill, size, shape, linetype, alpha, discrete ;
  • when using discrete, have to specify aesthetics

scale_colour_manual(..., values, aesthetics = "colour", breaks = waiver())

• ... Arguments passed on to discrete_scale, which is a discrete scale constructor.

  • palette A palette function that when called with a single integer argument (the number of levels in the scale) returns the values that they should take.
  • limits A character vector that defines possible values of the scale and their order
  • drop Should unused factor levels be omitted from the scale? The default, TRUE, uses the levels that appear in the data; FALSE to keep all the levels in the factor.
    • often set to drop=FALSE
  • na.value what aesthetic value should the missing values be displayed as? Useful to remove grey NA value area in figures.
    • often set to na.value = NA
  • breaks A character vector of breaks
  • labels A character vector giving scale labels (must be same length as breaks)
  • guide A function used to create a guide or its name.
    • guide = "legend" in scale_* is syntactic sugar for guide = guide_legend() (e.g. scale_color_manual(guide = "legend")). As for how to specify the guide for each scale in more detail, see guides().
    • Guides can be specified in each scale_* or in guides().
  • name The name of the scale. Used as the axis or legend title.

• values a set of aesthetic values to map data values to. The values will be matched in order (usually alphabetical) with the limits of the scale, or with breaks if provided. If this is a named vector, then the values will be matched based on the names instead. Data values that don’t match will be given na.value.

• aesthetics Character string or vector of character strings listing the name(s) of the aesthetic(s) that this scale works with. This can be useful, for example, to apply colour settings to the colour and fill aesthetics at the same time, via aesthetics = c("colour", "fill").

scale_fill_brewer(palette="Dark2") use brewer color palettes.

scale_fill_grey() Use grey scale

discrete_scale Discrete Scale Constructor.

discrete_scale(aesthetics, scale_name, palette, name = NULL, breaks = waiver(), labels = waiver(), legend = NULL, limits = NULL, expand = waiver(), na.value = NA, drop = TRUE, guide = "legend")

common discrete scale parameters: name, breaks, labels, na.value, limits and guide.

p <- ggplot(mtcars, aes(mpg, wt)) +
  geom_point(aes(colour = factor(cyl)))
p + scale_colour_manual(values = c("red", "blue", "green"))

# It's recommended to use a named vector
cols <- c("8" = "red", "4" = "blue", "6" = "darkgreen", "10" = "orange")
p + scale_colour_manual(values = cols)

# You can set color and fill aesthetics at the same time
ggplot(
  mtcars,
  aes(mpg, wt, colour = factor(cyl), fill = factor(cyl))
) +
  geom_point(shape = 21, alpha = 0.5, size = 2) +
  scale_colour_manual(
    values = cols,
    aesthetics = c("colour", "fill")
  )

# or you could just choose to plot a subset of category, ex. drop "10"
# As with other scales you can use breaks to control the appearance/order
# of the legend.
p + scale_colour_manual(
  # name = "Y series", # specify legend name here
  values = cols, 
  breaks = c("4", "6", "8"),
  labels = c("four", "six", "eight")
)


# And limits to control the possible values of the scale
p + scale_colour_manual(values = cols, limits = c("4", "8"))
p + scale_colour_manual(values = cols, limits = c("4", "6", "8", "10"))

Enlarge geom_point() dot size by setting size=3, default to 1.

# Set aesthetics to fixed value
ggplot(mtcars, aes(wt, mpg)) + 
    geom_point(colour = "red", size = 3)

ggplot(aes(x = Sepal.Length), data = iris) + 
  geom_histogram(color = 'black', fill = NA) + 
  geom_vline(aes(xintercept=median(iris$Sepal.Length),
                 color="median"), linetype="dashed",
             size=1) +
  geom_vline(aes(xintercept=mean(iris$Sepal.Length),
                 color="mean"), linetype="dashed",
             size=1) +
  scale_color_manual(name = "statistics", values = c(median = "blue", mean = "red"))

Reference lines: horizontal, vertical, and diagonal

geom_hline(slope, intercept), geom_vline(xintercept), geom_abline(yintercept) add reference lines (sometimes called rules) to a plot, either horizontal, vertical, or diagonal (specified by slope and intercept).

7.5.4 Add regression line

geom_smooth(mapping = NULL,data = NULL, stat = "smooth",position = "identity", ..., method = NULL,formula = NULL,se = TRUE, na.rm = FALSE,orientation = NA, show.legend = NA,inherit.aes = TRUE)

Addds a trend line over an existing plot. By default, it uses a LOESS smooth line. If you want a straight “linear model” line, you can use method=lm.

• method Smoothing method (function) to use, accepts either NULL or a character vector, e.g. "lm" (linear model), "glm", "gam", "loess" or a function, e.g. MASS::rlm or mgcv::gam, stats::lm, or stats::loess. "auto" is also accepted for backwards compatibility. It is equivalent to NULL.

  Defaults to loess (Locally Estimated Scatterplot Smoothing) when there are fewer than 1000 observations, and a gam when there are more observations. loess method stands for local regression fitting.

• formula Formula to use in smoothing function, eg. y ~ x, y ~ poly(x, 2), y ~ log(x). NULL by default, in which case method = NULL implies formula = y ~ x when there are fewer than 1,000 observations and formula = y ~ s(x, bs = "cs") otherwise.

• se=TRUE Whether to show the uncertainty band.

• level=.95 Level of confidence interval (CI) to use (0.95 by default).

• color color for the regression line

• fill color for the CI

• alpha transparency for the CI

• show.legend logical. Should this layer be included in the legends?

  • NA, the default, includes if any aesthetics are mapped. FALSE never includes, and
  • TRUE always includes. It can also be a named logical vector to finely select the aesthetics to display.

# fit a linear regression
geom_smooth(data=subset(long_format,key=="values_ma"),
                aes(color='lm1', linetype='lm1', size='lm1'), method=lm, se=FALSE) 

# fit a quadratic function
ggplot(CEF, aes(x=engine.size, y=mean_city.distance)) +
    geom_point() +
    geom_smooth(method='lm', formula = y ~ x + I(x^2), size = 1, se=FALSE) +
    theme(axis.text = element_text(size=rel(1.5)),
          axis.title = element_text(size=rel(1.5)))
# fit a degree 3 polynomial regression
p + geom_smooth(method = "lm", formula = y ~ poly(x, 3), se = FALSE)

Add regression equation to scatter plot

ggpubr::stat_regline_equation(label.y=NULL)

library(ggpubr) # add regression eq to figure
ggplot(df, aes_string(x="true value", y='predicted value')) +
    geom_point(shape=1) +    # Use hollow circles, default solid dot
    geom_text(aes(label=ifelse(Model %in% c('GFDL-ESM4', 'MIROC6'), as.character(Model),'')), hjust=-0.05, vjust=0, size=2) + # add text to outliers
    labs(subtitle="predicted v.s. true values", x='Simple global TCS [K]', y='Converted global TCS [K]') +
    geom_smooth(method=lm, se=TRUE) +
    geom_abline(intercept=0, slope=1, size=0.5, color='red', linetype="dashed") +
    stat_regline_equation(label.y=3.3) + # add regression equation
    theme_bw()

geom_text(data, mapping, check_overlap = FALSE, ...) add text to the plot

ggplot(mtcars, aes(wt, mpg)) +
    geom_text(aes(label = rownames(mtcars)))

ggrepel::geom_text_repel(aes(label=Model), size=3.5, fontface="bold") avoid overlap among labels.

geom_label() works similar to geom_text, except for that text is wrapped in a box.

Q: How to remove ‘a’ from legend when using aesthetics and geom_text?

A: Set show.legend = FALSE in geom_text.

legends for geom_text can only be called via color. If color is used before, then in order to keep the current color scheme, we have to add a new color scale, using ggnewscale::new_scale_color()+ and carrying on what you have to do afterwards.

https://stackoverflow.com/questions/59091627/add-new-legend-for-geom-text-with-text-labels-as-legend-key

library(ggplot2)
library(grid)
library(ggnewscale)

pfda_plot <- ggplot(data=pfdavar,aes(x=X1,y=X2,group=groups))+
  geom_point(aes(colour=groups))+
  geom_polygon(data=hulls,alpha=0.2,aes(fill=groups))+
  xlab("pFDA1")+
  ylab("pFDA2")+
  theme_classic()+
  theme(legend.title=element_blank())+
  new_scale_color()+   # define a ne color scheme
  geom_text(aes(label=labels,col=Species),
  fontface=1,hjust=0,vjust=0,size=3)+
  scale_color_manual(values=rep("black",18))

The above gives you something close, just that it is all ‘a’ for geom_text legend. What we need to do now, is change the default ‘a’, and for this I used @MarcoSandri’s solution to change the default “a” in legend for geom_text()

g <- ggplotGrob(pfda_plot)
lbls <- 1:18
idx <- which(sapply(g$grobs[[15]][[1]][[1]]$grobs,function(i){
  "label" %in% names(i)}))
for(i in 1:length(idx)){
g$grobs[[15]][[1]][[1]]$grobs[[idx[i]]]$label <- lbls[i]
}
grid.draw(g)

geom(text)

With geom_text or annotate in ggplot2, you can set a number of properties of the text. geom_text is used to add text from the data frame, and annotate is used to add a single text element.

Name	Default value
size	5
family	"" (sans)
fontface	plain
lineheight	1.2
angle	0
hjust	0.5
vjust	0.5

---

7.5.5 Subplots

Note that the group must be called in the X argument of ggplot(aes(x = group, fill = subgroup)).

The subgroup is called in the fill argument.

The facet_wrap(~class, nrow = NULL, ncol = NULL, scales = "fixed") ( Multiple plots by factor in ggplot (facets) ) gives out each variable in an individual panel grouped by class. ~class can also be vars(class).

• facet_grid() function will produce a grid of plots for each combination of variables that you specify, even if some plots are empty.
• scales='fixed' if subplots share x-axes or y-axes; scales='free' for each plot having its own axes.
• nrow, ncol define #of rows/cols

geom_errorbar() A geom that draws error bars, defined by an upper and lower value. This is useful e.g., to draw confidence intervals.

7.5.6 Parameters

• x - (required) x coordinate of the bar
• ymin - (required) y coordinate of the lower whisker
• ymax - (required) y coordinate of the upper whisker
• size - (default: 0.5) thickness of the lines
• linetype - (default: 1=solid) the type of the lines
• colour - (default: “black”) the color of the lines
• width - (default: 0.9) width of the whiskers
• alpha - (default: 1=opaque) the transparency of the lines

plot_data <-  Rad_trend_decade_allCON %>% 
    gather(seasons, values, -CON) %>% 
    mutate(seasons = factor(seasons, levels=c("DJF", "MAM", "JJA", "SON", "ANN") ),
           CON = factor(CON, levels=c(CON_levels, "WD") ) 
          )
plot_data


## group by season
p_season_box <- ggplot(plot_data, aes(x=CON, y=values, fill=CON) ) +
    stat_boxplot(geom ='errorbar',
                 position = position_dodge(width = 0.9)
                ) +
    geom_boxplot(outlier.shape = NA,
                 position = position_dodge(width = 0.9)
                ) +
    scale_y_continuous(limits = c(-7, 7), # set limits of y-axis
                       breaks = seq(-7, 7, by=2)
                      ) + 
        facet_wrap(~seasons, nrow=1)

p_season_box 

Ignore outliers

Sometimes it can be useful to hide the outliers, for example when overlaying the raw data points on top of the boxplot. Hiding the outliers can be achieved by setting outlier.shape = NA. But the outliers would still affect the y-axis scale and make your box condensed. You need to set outliers = FALSE too.

# set y-axis limits mannually
ggplot(the_variable, aes(x=Water_receive, y=water_stress, fill=year)) +
    geom_boxplot(outlier.shape = NA) +
    scale_y_continuous(limits = quantile(the_variable$water_stress, c(0.1, 0.9)))

# set outliers = FALSE by discarding outliers from the plot
ggplot(the_variable, aes(x=Water_receive, y=water_stress, fill=year)) +
    geom_boxplot(outlier.shape = NA, outliers = FALSE) 

Deal with Outliers

One idea would be to winsorize the data in a two-pass procedure:

1. run a first pass, learn what the bounds are, e.g. cut of at given percentile, or N standard deviation above the mean, or …
2. in a second pass, set the values beyond the given bound to the value of that bound

I should stress that this is an old-fashioned method which ought to be dominated by more modern robust techniques but you still come across it a lot.

Grouped bar plot

The items on the x-axis have x values of 1, 2, 3, and so on, though you typically don’t refer to them by these numerical values.

• When you use geom_bar(width = 0.9), it makes each group take up a total width of 0.9 on the x-axis. 组间宽度。
• When you use position_dodge(width = 0.9), it spaces the bars so that the middle of each bar is right where it would be if the bar width were 0.9 and the bars were touching. 值越大，同一组的bar之间越远。组内bar之间的间距。

Another option is to calculate stats first, and plot geom_crossbar. It is much faster this way.

Rad_trend_decade_group <- Rad_trend_decade_allCON %>% group_by(CON)
groups <- Rad_trend_decade_group %>% 
    group_split()
group_keys(Rad_trend_decade_group)

groups[[1]] %>% select(-CON) %>% apply(2, function(group) boxplot.stats(group)$stats)

trend_summary_tibble <- Rad_trend_decade_group %>% 
    group_modify(~{
        .x %>% 
            select(c("DJF", "MAM", "JJA", "SON", "ANN")) %>% 
            apply(2, function(group) 
                boxplot.stats(group)$stats) %>% 
            as_tibble()
    }) %>% ungroup()
trend_summary_tibble <- trend_summary_tibble %>% 
      mutate(stats = rep(c("low.whisker", "1st.Q", "median", "3rd.Q", "upper.whisker"), 7 ) )
trend_summary_tibble
# A tibble: 35 x 7
# CON      DJF    MAM     JJA    SON    ANN stats        
# <chr>  <dbl>  <dbl>   <dbl>  <dbl>  <dbl> <chr>        
# 1 AF    -1.43  -1.44  -2.43   -1.29  -0.768 low.whisker  
# 2 AF     0.428  0.506 -0.160   0.423  0.543 1st.Q        
# 3 AF     1.02   1.11   0.393   0.966  0.988 median       
# 4 AF     1.68   1.84   1.41    1.58   1.42  3rd.Q        
# 5 AF     3.54   3.84   3.77    3.32   2.73  upper.whisker
# 6 AS    -1.72  -4.17  -6.84   -3.45  -2.96  low.whisker  
# 7 AS    -0.282 -0.774 -1.65   -0.974 -0.652 1st.Q        
# 8 AS     0.192  0.296  0.0631 -0.116  0.168 median       
# 9 AS     0.673  1.49   1.81    0.674  0.888 3rd.Q        
# 10 AS     2.11   4.89   7.00    3.15   3.18  upper.whisker

CON_levels <- trend_summary_tibble %>% 
    filter((stats=="median") &(CON!="WD") ) %>% 
    arrange(ANN) %>% pull(CON)
CON_levels

plot_data <- trend_summary_tibble %>% 
    gather(seasons, values, -CON, -stats) %>% 
    spread(stats, values)
plot_data <- plot_data %>% 
    mutate(seasons = factor(seasons, levels=c("DJF", "MAM", "JJA", "SON", "ANN") ),
           CON = factor(CON, levels=c(CON_levels, "WD") )
    )
plot_data
# A tibble: 9,072,000 x 3
#  CON   seasons   values
# <fct>   <fct>     <dbl>
# 1 EU    DJF         NA
# 2 EU    DJF         NA
# 3 EU    DJF         NA
# 4 EU    DJF         NA
# 5 EU    DJF         NA      
      
x11()
dodge <- position_dodge(width=0.9)
## Group by continent
# dev.set(dev.prev())
p_box <- ggplot(plot_data, aes(x=seasons, y=median, fill = seasons) ) +
    geom_crossbar(aes(ymin = `1st.Q`, ymax = `3rd.Q`), width = 0.8, size=0.35, position = dodge) +
    geom_errorbar(aes(ymin = low.whisker, ymax = upper.whisker, ), width = 0.8, position = dodge) + 
    facet_wrap(~CON, nrow=1) + # subplots group by 'CON'
    labs(y=TeX("SSR Trend \\[$Wm^{-2}/dec$\\]")) +
    scale_y_continuous(breaks=seq(-6,6,2)) +
    guides(fill = guide_legend(title="Seasons") ) + # change legend title
    theme(strip.text = element_text(size=10, face="bold"),
          axis.text.x = element_text(size=10, face="bold", angle = 90),
          axis.text.y = element_text(size=10, face="bold"),
          axis.title.x = element_blank(),
          legend.position = "bottom",
          legend.title = element_text(size=10, face="bold"),
          legend.text = element_text(face="bold")
          )
p_box      

7.5.7 guides(), together with guide_legend()

https://ggplot2-book.org/guides.html#sub-layers-legends

Guides for each scale can be set scale-by-scale with the guide argument in scale_*_manual(), or en masse with guides(...).

• ... List of scale name-guide pairs. The guide can either be
  • a string (i.e. “colorbar” or “legend”), or
  • a call to a guide function (i.e. guide_colourbar() or guide_legend() ) specifying additional arguments.

# ggplot object

dat <- data.frame(x = 1:5, y = 1:5, p = 1:5, q = factor(1:5),
 r = factor(1:5))
p <- ggplot(dat, aes(x, y, colour = p, size = q, shape = r)) + geom_point()

# without guide specification
p

# Show colorbar guide for colour.
# All these examples below have a same effect.

p + guides(colour = "colorbar", size = "legend", shape = "legend")
p + guides(colour = guide_colorbar(), size = guide_legend(),
  shape = guide_legend())
p +
 scale_colour_continuous(guide = "colorbar") +
 scale_size_discrete(guide = "legend") +
 scale_shape(guide = "legend")

 # Remove some guides
 p + guides(colour = "none")
 p + guides(colour = "colorbar", size = "none")

##
# Guides are `integrated` where possible
##
p + guides(colour = guide_legend("title"), size = guide_legend("title"),
  shape = guide_legend("title"))

# same as (more concise)
g <- guide_legend("title")
p + guides(colour = g, size = g, shape = g)

p + theme(legend.position = "bottom")

# Set order for multiple guides/legends #
ggplot(mpg, aes(displ, cty)) +
  geom_point(aes(size = hwy, colour = cyl, shape = drv)) +
  guides(
   colour = guide_colourbar(order = 1),
   shape = guide_legend(order = 2),
   size = guide_legend(order = 3)
 )

guide_legend(title = waiver(), label = TRUE, keywidth = NULL, keyheight = NULL, override.aes = list(), nrow = NULL, ncol = NULL, byrow = FALSE, reverse = FALSE,order = 0, ...) Legend type guide shows key (i.e., geoms) mapped onto values. Legend guides for various scales are integrated if possible.

• title A character string or expression indicating a title of guide.
  • NULL, the title is not shown.
  • By default (waiver()), the name of the scale object or the name specified in labs() is used for the title.
• override.aes Takes a list of aesthetic parameters that will override the default legend appearance.
• nrow, ncol The desired number of rows/columns of legends.
• reverse logical. If TRUE the order of legends is reversed.
• order positive integer less than 99 that specifies the order of this guide among multiple guides. This controls the order in which multiple guides are displayed, not the contents of the guide itself.

7.5.8 Template for multi-series

Multiple groups aesthetics, e.g., color, size, linetype, …, if you want an integrated legend, need to set the same title for the aesthetics.

g <- guide_legend("title", nrow=2, byrow=TRUE) # define legend aes
# colors
cols <- c("values"="blue", "values_ma"="blue", "lm1"="red", 
          "values2"="black", "values_ma2"="black", "lm2"="red")
# if you don't want to repeat the breaks, you may use `setNames`
# cols <- setNames(ob = c("blue", "blue", "red", "black", "black", "red"),
#                        nm = c("values", "values_ma", "lm1", "values2", "values_ma2", "lm2"))

# line widths
sizes <- c("values"=0.3, "values_ma"=0.7, "lm1"=0.7,
           "values2"=0.3, "values_ma2"=0.7, "lm2"=0.7)
# linetypes
lines <- c("values"="dashed", "values_ma"="solid", "lm1"="solid",
           "values2"="dashed", "values_ma2"="solid", "lm2"="solid")
# breaks:
#           — if omit a series in `breaks`, the series will be droped; 
#           — rearrange legend order; 
breaks <- c("values", "values_ma", "lm1", 
            "values2", "values_ma2", "lm2")
# labels, must match the length of `breaks`
labels <- c("raw", "Gaussian MA", 
            unname(TeX( sprintf("Reg Line \\textbf{1961-2019}, \\[%.2f, %.2f, %.2f\\] \\[$Wm^{-2}/dec$\\]", decadal_trend["beta"], decadal_trend["lower bound"], decadal_trend["upper bound"]) ) ),
            "raw2", "Gaussian MA2", 
            unname(TeX( sprintf("Reg Line2 \\textbf{%s-2019}, \\[%.2f, %.2f, %.2f\\] \\[$Wm^{-2}/dec$\\]", bk_point+1, decadal_trend_2["beta"], decadal_trend_2["lower bound"], decadal_trend_2["upper bound"]) ) )
)

ggplot() + scale_colour_manual(values = cols, breaks=breaks, labels = labels) +
    scale_linetype_manual(values=lines, breaks=breaks, labels = labels) +
    scale_size_manual(values=sizes, breaks=breaks, labels = labels) +
    theme_bw() +
    guides(colour=g, linetype=g, size=g # specify all aes for legend at once
    ) +
    theme(title = element_text(size=10), # title size
          axis.title.x = element_blank(), # axis lable
          legend.title = element_blank(), # remove legend title
          legend.text = element_text(size=8), # legend scale size
          legend.position = c(0.37, 0.9), 
          legend.direction = "horizontal",
          legend.spacing.y = unit(0, 'mm'), # spacing between legend title and legends;
          legend.text.align = 0, # legend key align to the left
          legend.key.height = unit(0.8,"line"), # spacing between legend items;
          legend.margin = margin(t=0, b=0, unit="mm") # margins around legend box
    )

• legend.position Can be text placement. Five possible values: “left”, “top”, “right”, “bottom”, or "none" (disable legend).
  • The argument legend.position can be also a numeric vector c(x,y). In this case it is possible to position the legend inside the plotting area. x and y are the coordinates of the legend box. Their values should be between 0 and 1. c(0,0) corresponds to the “bottom left” and c(1,1) corresponds to the “top right” position.
• legend.key.height the height of the legend key; reduce or add vertical spacing between legend items.
• legend.key.width the width of the legend key;
• legend.text.align = 0 align legend keys/text to the left;
• legend.box.background = element_rect(colour = "black", size=1) add box to legend
• legend.direction layout of items in legends (“horizontal” or “vertical”)
• legend.box arrangement of multiple legends (“horizontal” or “vertical”)

guides() re-define guides, an example for override.aes; could also be used to remove some legends by specifying e.g., color="none".

guides(fill = FALSE,
       color = guide_legend(override.aes =
                      list(fill = c("converted"="#00BFC4", "reported"="#FF6666", obs_col=NA),
                           color = c("converted"="#00BFC4", "reported"="#FF6666", obs_col="#010912")
                           )
                      )
         )

guides(linetype = FALSE,
       size = FALSE,
       color = guide_legend(
           override.aes = list(
               linetype = lines,
               color = cols,
               size = sizes 
                 ),
           nrow = 2,
           byrow = TRUE
           )
      )

Change legend title

Specify a new legend title for the color aesthetic:

• labs(color="My new title")
• guides(color=guide_legend("My new title"))

Putting two different legends in two columns

theme(legend.box = "horizontal")

Add legend to geom_vline

Need to specify color inside aes, then use scale_color_manual to specify legends.

ggplot(end_climate, aes(x=end.pre)) +
    geom_histogram(aes(y = after_stat(density)), binwidth=0.1, fill="#BDBCBC", color="black") + 
    geom_vline(aes(xintercept=1.734513, color="Burke"), linetype="dashed") +
    scale_color_manual(values=c("Burke"="red"), labels=c("Burke"="Burke's fixed optimum")) +
    labs(x="Precipitation in 2019 [Meters]") +
    theme(legend.position = c(0.8, 0.9),
          legend.title = element_blank(),
          axis.title.y = element_blank(), )

legend alignment (irregular legends)

split legends into multiple ragged rows/columns (with different length) as you desire

https://stackoverflow.com/questions/27803710/ggplot2-divide-legend-into-two-columns-each-with-its-own-title

Key idea:

1. create a dummy factor level and setting its colour to white in the legend, so that it can’t be seen.
2. so now we have regular levels. Then, we use scale_fill_manual to set the color of this blank level to “white”. drop=FALSE forces ggplot to keep the blank level in the legend.
3. Remember to factor the category column.

## `factor` the category column, and specify `drop=FALSE`
diamonds$cut = factor(diamonds$cut, levels=c("Fair","Good"," ","Very Good",
                                             "Premium","Ideal"))

ggplot(diamonds, aes(color, fill=cut)) + geom_bar() + 
  scale_fill_manual(values=c(hcl(seq(15,325,length.out=5), 100, 65)[1:2], 
                             "white",
                             hcl(seq(15,325,length.out=5), 100, 65)[3:5]),
                    drop=FALSE) +
  guides(fill=guide_legend(ncol=2)) +
  theme(legend.position="bottom")

Retrieve environment variable

Sys.getenv('PLOTLY_MATHJAX_PATH')
# set an environment variable
Sys.setenv('PLOTLY_MATHJAX_PATH' = '/Users/Menghan/Documents/R/MathJax')

Check global setting/options

options(...) set options, using name = value.

getOption(x, default=NULL) get option values.

• default if the specified option is not set in the options list, this value is returned. This facilitates retrieving an option and checking whether it is set and setting it separately if not.

> options("device")
$device
[1] "RStudioGD"

Check package version

packageVersion("snow")

Length Unit

The relation between the absolute units is as follows: 1in = 2.54cm = 25.4mm = 72pt = 6pc

Latex symbol in legend labels

cols <- c() # specify col vector here
sizes <- c() # specify linewidth 
lines <- c() # linetype
breaks <- c() # group breaks, set legend order;

## note that the use of `unname()` is necessary
labels <- unname(latex2exp::TeX(c("$A_{t-k}^h$", "$B_{t-k}^h$")))

scale_colour_manual(values = cols, breaks=breaks, labels = labels) +
    scale_linetype_manual(values=lines, breaks=breaks, labels = labels) +
    scale_size_manual(values=sizes, breaks=breaks, labels = labels) 

7.5.9 grid.gedit()

https://bookdown.org/rdpeng/RProgDA/the-grid-package.html#grobs

https://stackoverflow.com/questions/15059093/ggplot2-adjust-the-symbol-size-in-legends

grid.gedit(..., grep = TRUE, global = TRUE) Changes the value of one of the slots of a grob and redraws the grob.

• ... Zero or more named arguments specifying new slot values.

# To get the names of all the grobs in the ggplot
grid.ls(grid.force())    

# The edit - to set the size of the point in the legend to 4 mm
grid.gedit("key-[-0-9]-1-1", size = unit(4, "mm"))    

Transparency in Rstudio

https://tinyheero.github.io/2015/09/15/semi-transparency-r.html

Error Message:

Warning message:
In grid.Call.graphics(L_polygon, x$x, x$y, index) :
  semi-transparency is not supported on this device: reported only once per page

Fix:

I was able to solve this issue by switching over to use the Cairo graphics device. Make sure you first install the Cairo R package.

install.packages("Cairo")

Once you have that installed, place the following in your ~/.Rprofile

setHook(packageEvent("grDevices", "onLoad"),
function(...) grDevices::X11.options(type='cairo'))
options(device='x11') # set default Graphic Device

This makes it so that your default graphics device is set to Cairo whenever you start a new R session. Now open a new R session, and try the same plotting code (from above). You should see a plot with transparency.

# get default graphic device
getOption("device")

What’s a graphic device?

It’s the engine that renders your plot. Common graphics devices are Quartz and X11.