2. ggplot2

The ggplot2 package helps you plot graphics with a grammar. The grammar provides a way to talk about parts of a plot. The grammar talks about the following components of a plot.

  • data is what is being plotted

  • geometric objects are the shapes and lines that appear on the plot

  • aesthetics are the appearance of the geometric objects and the mapping of variables to such aesthetics

  • position adjustment is the placement of elements

  • scale is the range of values for each aesthetic mapping

  • coordinate system is used to organize the geometric objects

  • facets are groupos of data shown in differen plots

2.1. Geometries

2.1.1. geom_point

The geom_point is used for drawing individual points.

[1]:

library('ggplot2')
library('repr')

df <- data.frame(
    x1 = rnorm(1000),
    x2 = rnorm(1000)
)

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2))
_images/ggplot2_3_0.png

You may also style with the color attribute.

[2]:

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2), color='firebrick')
_images/ggplot2_5_0.png

You may also style with themes.

[3]:

library('ggthemes')

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2), color='firebrick') +
    theme_economist()
_images/ggplot2_7_0.png

2.1.2. geom_hex

The geom_hex function is used for drawing individual points as hexagons.

[4]:

df <- data.frame(
    x1 = rnorm(1000),
    x2 = rnorm(1000)
)

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_hex(mapping=aes(x=x1, y=x2))
_images/ggplot2_9_0.png

2.1.3. geom_line

The geom_line function is used for drawing lines.

[5]:

df <- data.frame(
    x1 = seq(1, 1000),
    x2 = seq(1, 1000)
)

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_line(mapping=aes(x=x1, y=x2))
_images/ggplot2_11_0.png

2.1.4. geom_smooth

The geom_smooth function is used to draw a smoothed line.

[6]:

df <- data.frame(
    x1 = seq(1, 1000),
    x2 = seq(1, 1000) + (rnorm(1000) * seq(1, 1000))
)

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2)) +
    geom_smooth(mapping=aes(x=x1, y=x2))

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
_images/ggplot2_13_1.png

2.1.5. geom_col

The geom_col function is used to draw bars.

[7]:

df <- data.frame(
    x1 = c('DC', 'MD', 'VA'),
    x2 = c(20, 30, 15)
)

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_col(mapping=aes(x=x1, y=x2))
_images/ggplot2_15_0.png

2.1.6. geom_boxplot

Use geom_boxplot to plot box-whisker plots.

[8]:

df <- data.frame(
    x1 = c(rnorm(100, mean=15, sd=5), rnorm(100, mean=25, sd=5)),
    x2 = c(rnorm(100, mean=15, sd=5), rnorm(100, mean=25, sd=5)),
    state = c(rep('DC', 100), rep('VA', 100))
)

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_boxplot(mapping=aes(x=state, y=x1))
_images/ggplot2_17_0.png

You may play around with attributes to change the look and feel of the box-whisker plot.

[9]:

options(repr.plot.width=5, repr.plot.height=3)

ggplot(df, mapping=aes(x=state, y=x1, color=state)) +
    geom_boxplot(
                 outlier.colour='green',
                 outlier.shape=8,
                 notch=TRUE
                ) +
    coord_flip() +
    geom_jitter(alpha=0.5, position=position_jitter(width=0.2))
_images/ggplot2_19_0.png

2.1.7. geom_tile

Use geom_tile to build correlation matrix plots.

[10]:

df <- data.frame(
    v1 = c('a', 'a', 'a', 'b', 'b', 'c'),
    v2 = c('a', 'b', 'c', 'b', 'c', 'c'),
    corr = c(1.0, -0.2, 0.2, 1.0, -1.0, 1.0)
)

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df, mapping=aes(x=v2, y=v1)) +
    geom_tile(data=df, aes(fill=corr), color='white') +
    scale_fill_gradient2(low='blue', high='red', mid='white', midpoint=0, limit=c(-1, 1)) +
    theme(axis.text.x=element_text(angle=45, vjust=1, size=11, hjust=1)) +
    coord_equal()
_images/ggplot2_21_0.png

2.2. Aesthetic mapping

Aesthetics may be used to change colors.

[11]:

df <- data.frame(
    x1 = c(rnorm(100, mean=1, sd=1), rnorm(100, mean=2.5, sd=1)),
    x2 = c(rnorm(100, mean=1, sd=1), rnorm(100, mean=2.5, sd=1)),
    state = c(rep('DC', 100), rep('VA', 100))
)

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8)
_images/ggplot2_23_0.png

2.3. Position

Positioning can change the placement of elements and produce different types of plots.

[12]:

df <- data.frame(
    x1 = c('DC', 'MD', 'VA'),
    x2 = c(20, 30, 15)
)

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_col(mapping=aes(x=x1, y=x2, color=x1, fill=x1))
_images/ggplot2_25_0.png 
[13]:

library('tidyr')

df <- data.frame(
    state = c('DC', 'MD', 'VA'),
    water = c(10, 20, 30),
    energy = c(12, 25, 39),
    social = c(15, 20, 10)
)

n <- df %>%
        pivot_longer(-state, names_to='expense', values_to='cost')

options(repr.plot.width=5, repr.plot.height=4)

ggplot(n) +
    geom_col(mapping=aes(x=state, y=cost, fill=expense))
_images/ggplot2_26_0.png

When position='fill', the stacked bars are forced to 100%.

[14]:

options(repr.plot.width=5, repr.plot.height=4)

ggplot(n) +
    geom_col(mapping=aes(x=state, y=cost, fill=expense), position='fill')
_images/ggplot2_28_0.png

When position='dodge the bars are side-by-side.

[15]:

options(repr.plot.width=5, repr.plot.height=4)

ggplot(n) +
    geom_col(mapping=aes(x=state, y=cost, fill=expense), position='dodge')
_images/ggplot2_30_0.png

2.4. Scale

Scaling may help to zoom in or out of the plot, as well as rescale the axes.

[16]:

df <- data.frame(
    x1 = c(rnorm(100, mean=10, sd=1), rnorm(100, mean=25, sd=1)),
    x2 = c(rnorm(100, mean=10, sd=1), rnorm(100, mean=25, sd=1)),
    state = c(rep('DC', 100), rep('VA', 100))
)

[17]:

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8) +
    scale_color_brewer(palette='Set3') +
    scale_x_continuous() +
    scale_y_continuous()
_images/ggplot2_33_0.png 
[18]:

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8) +
    scale_color_brewer(palette='Set3') +
    scale_x_continuous(limits=c(0, 15)) +
    scale_y_continuous(limits=c(0, 15)) +
    theme(legend.position='none')

Warning message:
“Removed 100 rows containing missing values (geom_point).”
_images/ggplot2_34_1.png 
[19]:

options(repr.plot.width=4, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8) +
    scale_color_brewer(palette='Set3') +
    scale_x_continuous(limits=c(20, 30)) +
    scale_y_continuous(limits=c(20, 30)) +
    theme(legend.position='none')

Warning message:
“Removed 100 rows containing missing values (geom_point).”
_images/ggplot2_35_1.png 
[20]:

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8) +
    scale_color_brewer(palette='Set3') +
    scale_x_log10() +
    scale_y_log10()
_images/ggplot2_36_0.png

2.5. Coordinate

Coordinate functions such as coord_flip and coord_polar can change the type and look of a plot.

[21]:

df <- data.frame(
    state = c('DC', 'MD', 'VA'),
    water = c(10, 20, 30),
    energy = c(12, 25, 39),
    social = c(15, 20, 10)
)

n <- df %>%
        pivot_longer(-state, names_to='expense', values_to='cost')

options(repr.plot.width=5, repr.plot.height=4)

ggplot(n) +
    geom_col(mapping=aes(x=state, y=cost, fill=expense), position='dodge') +
    coord_flip()
_images/ggplot2_38_0.png 
[22]:

options(repr.plot.width=5, repr.plot.height=4)

ggplot(n) +
    geom_col(mapping=aes(x=state, y=cost, fill=expense), position='dodge') +
    coord_polar()
_images/ggplot2_39_0.png

2.6. Facets

Facets can create subplots.

[23]:

df <- data.frame(
    state = c('DC', 'MD', 'VA'),
    water = c(10, 20, 30),
    energy = c(12, 25, 39),
    social = c(15, 20, 10)
)

n <- df %>%
        pivot_longer(-state, names_to='expense', values_to='cost')

options(repr.plot.width=7, repr.plot.height=3)

ggplot(n) +
    geom_col(mapping=aes(x=expense, y=cost, fill=expense)) +
    facet_wrap(~state) +
    theme(legend.position='none')
_images/ggplot2_41_0.png

2.7. Labels and annotations

You may modify the title, axes and legend with labs. You may use geom_label_repel from the ggrepel library to annotate specific coordinates.

[24]:

library('ggrepel')

df <- data.frame(
    x1 = c(rnorm(100, mean=1, sd=1), rnorm(100, mean=2.5, sd=1)),
    x2 = c(rnorm(100, mean=1, sd=1), rnorm(100, mean=2.5, sd=1)),
    state = c(rep('DC', 100), rep('VA', 100))
)

centers <- data.frame(
    x1 = c(1.0, 2.5),
    x2 = c(1.0, 2.5),
    state = c('DC', 'VA')
)

options(repr.plot.width=5, repr.plot.height=4)

ggplot(df) +
    geom_point(mapping=aes(x=x1, y=x2, color=state), alpha=0.8) +
    geom_label_repel(data=centers, mapping=aes(x=x1, y=x2, label=state)) +
    labs(
        title ='the-title',
        x='x-axis',
        y='y-axis',
        color='the-legend'
    ) +
    theme(
        plot.title=element_text(
            size=15,
            face='bold',
            margin=margin(10, 0, 10, 0),
            vjust=1,
            family='Times'
        ),
        plot.background=element_rect(fill='grey')
    )
_images/ggplot2_43_0.png

2.8. Choropleth map

[25]:

suppressMessages({
    library('dplyr')
})

allStates <- unique(map_data('state')$region)
x <- rnorm(length(allStates), mean=1, sd=1)

randomData <- data.frame(
    region=allStates,
    weight=x,
    stringsAsFactors=FALSE
)

df <- randomData %>%
    left_join(map_data('state'), by='region')

ggplot(df) +
    geom_polygon(
        mapping=aes(x=long, y=lat, group=group, fill=weight),
        color='white',
        size=0.1
    ) +
    coord_map() +
    scale_fill_continuous(low='#132B43', high='Red')
_images/ggplot2_45_0.png

2.9. Subplots

Use the grid package.

[26]:

suppressMessages({
    library('grid')
    library('gridExtra')
})


allStates <- unique(map_data('state')$region)

randomData1 <- data.frame(
    region=allStates,
    weight=rnorm(length(allStates), mean=1, sd=1),
    stringsAsFactors=FALSE
)

randomData2 <- data.frame(
    region=allStates,
    weight=rnorm(length(allStates), mean=10, sd=2),
    stringsAsFactors=FALSE
)

df1 <- randomData1 %>%
    left_join(map_data('state'), by='region')
df2 <- randomData2 %>%
    left_join(map_data('state'), by='region')

plt1 <- ggplot(df1) +
            geom_polygon(
                mapping=aes(x=long, y=lat, group=group, fill=weight),
                color='white',
                size=0.1
            ) +
            coord_map() +
            scale_fill_continuous(low='#132B43', high='Red') +
            theme(legend.position='none')

plt2 <- ggplot(df2) +
            geom_polygon(
                mapping=aes(x=long, y=lat, group=group, fill=weight),
                color='white',
                size=0.1
            ) +
            coord_map() +
            scale_fill_continuous(low='#132B43', high='Blue') +
            theme(legend.position='none')

options(repr.plot.width=10, repr.plot.height=4)

grid.arrange(plt1, plt2, ncol=2)
_images/ggplot2_47_0.png