Skip to content

Latest commit

 

History

History
726 lines (630 loc) · 23.6 KB

README.md

File metadata and controls

726 lines (630 loc) · 23.6 KB
Raw

vagotheme

The main purpose of vagotheme is to enable R users to use the colours and styles defined in the Victoria Auditor General’s Office’s style guide to create compliant visualisations in ggplot.

This package is not endorsed by VAGO.

vagotheme was created with the palettes package, which provides a comprehensive library for colour vectors and colour palettes using a new family of colour classes (palettes_colour, and palettes_palette) that always print as hex codes with colour previews. Colour palette packages created with palettes have access to the following capabilities, all without requiring you to write any code: formatting, casting and coercion, extraction and updating of components, plotting, colour mixing arithmetic, and colour interpolation.

See the following vignettes to learn how to use palletes with other packages: - Using palettes with ggplot2 - Using palettes with gt - Using palettes with biscale - Compatibility with other colour packages

Installation

You can install the development version of vagotheme from GitHub with:

# install.packages("devtools")
devtools::install_github("zerogetsamgow/vagotheme")

Included Colours

Base colours

vagotheme’s base colour palette vago_colours$base comprises 7 colours.

Theme colours

vago’s theme colours vago_colours$theme reorders the colours to match the order in the Windows theme.

Chart

vago’s chart colours vago_colours$chartis a palette of four colours that is the primary palette for charts .

Other

An additional non palette colours vago.lilac is included to use as a background colour.

Fonts

vagotheme uses the Noto Sans font from sysfonts::font_add_google() instead of the Windows Font Segoe UI to enable easier installation.

Installation

You can install the development version of vagotheme from GitHub with:

# install.packages("devtools")
devtools::install_github("zerogetsamgow/vagotheme")

Usage

{vagotheme} is designed to produce {gglot2} that comply with the vago style guide simply. For example.

## basic example code
ggplot2::ggplot(data=iris,aes(x=Sepal.Length, y = Petal.Length,colour=Species)) +
  geom_point(size = 3) +
  scale_colour_manual(values=(vago_colours$chart),labels=stringr::str_to_title) +
  scale_x_continuous(name="Sepal length")+
  scale_y_continuous(name="Petal length")+
  theme_vago_white()

{vagotheme} exports two themes - theme_vago_white()(seen above) and theme_vago_lilac() to enable plots to be produced for any VAGO publication.

## example plot
ggplot(data=iris,aes(x=Sepal.Length, y = Petal.Length,colour=Species)) +
  geom_point(size=3) +
  scale_colour_manual(values=vago_colours$theme,labels=stringr::str_to_title) +
  scale_x_continuous(name="Sepal length")+
  scale_y_continuous(name="Petal length")+
  labs(title="Sepal and petal lengths of irises")+
  theme_vago_lilac()

As well as utilising vago colour these themes are rendered using the Segoe UI font.

VAGO Charts

Column charts

Auditors use column charts most often because they are useful for displaying and comparing discrete data sets, which is what we encounter most. The order of the bars depends on the categories they depict. • When there is a natural sequence we typically order the column left to right (for example, a series of dates may be ordered earliest to most recent). • If the dataset is not contiuous (eg. categorical or nominal comparisons), we advise to use horizontal bar charts instead. • Consider removing axis titles if they are self-explanatory, or fully explained in the heading.

# Create tibble of data for examples
car_accidents = 
  tibble(
    year = c(2020,2021,2022,2023,2024),
    accidents = c(10,   5,  45, 10, 16)
  )

## basic example of a column plot
ggplot(data=car_accidents,aes(x=year, y = accidents)) +
  geom_col(fill = vago.blue) +
  geom_text(aes(label = accidents), size = 5, nudge_y = 1.5)+
  scale_x_continuous(name = NULL, breaks = 2020:2024)+
  scale_y_continuous(name = NULL, labels = NULL,  expand = c(0,0,.1,.1))+
  labs(title="Simple column charts",
       subtitle = 
         str_wrap(
           "One-dimensional column chart: one colour, single axis, y-axis can be replaced by data label",
       70)
       )+
  theme_vago_white()

# Create tibble of data for examples
car_accidents = 
  tibble(
    month = rep(c("Jan","Feb","Mar","Apr","May"),4),
    freeway = c(rep("M1",5),rep("M2",5),rep("M3",5),rep("M4",5)),
    accidents = c(10, 5, 45, 10,    16, 
                  20, 10, 40, 12, 10, 
                  30, 15, 25, 14, 18, 
                  25, 20, 15, 18, 10)
  ) |> 
  mutate(month = factor(month, levels = month.abb),
         freeny = factor(freeway))

## example of multi colour column plot
ggplot(data=car_accidents,aes(x=month, y = accidents, fill = freeway)) +
  geom_col(position = position_dodge2()) +
  scale_x_discrete(name = NULL)+
  scale_y_continuous(name = NULL, expand = c(0,0))+
  scale_fill_manual(name = NULL, values = vago_colours$chart)+
  labs(title="Two-dimensional histogram: full colour",
       subtitle =
         str_wrap(
           "If you have more than 4 categories, please consider using different method to present.",
           70))+
  theme_vago_white()

## example of single colour column plot
ggplot(
  data=car_accidents,
  aes(x=month, y = accidents, fill = freeway)) +
  geom_col(position = position_dodge2()) +
  scale_x_discrete(name = NULL)+
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,51))+
  scale_fill_manual(name = NULL, values = c(rep(vago.grey,3), vago.blue))+
  labs(title="Two-dimensional histogram: single series highlight",
       subtitle = str_wrap(
         "If you want to highlight one category and de-emphasise the remaining ones",
         70)
       )+
  theme_vago_white()

## example of multi colour stacked column plot
ggplot(
  data = car_accidents, 
  aes(
    x = month, 
    y = accidents, 
    fill = fct_rev(freeway), 
    label = accidents), 
    colour = fct_rev(freeway)) +
  geom_col(position = position_stack()) +
  geom_text(aes(colour= fct_rev(freeway)), position = position_stack(vjust = .5))+
  scale_x_discrete(name = NULL)+
  guides(fill = guide_legend(reverse = TRUE)) +
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,131))+
  scale_fill_manual(name = NULL, values = rev(vago_colours$chart))+
  scale_colour_manual(guide = "none", name = NULL, values = c("black","black","black","white"))+
  labs(title="Stacked column chart: full colour",
       subtitle = 
         str_wrap(
           "Effective for illustrating how individual parts contribute to a whole.",
           70)
       )+
  theme_vago_white()

## example of single colour column plot
ggplot(data=car_accidents, 
       aes(x=month, 
           y = accidents, 
           fill = fct_rev(freeway), 
           label = accidents), 
           colour = fct_rev(freeway)) +
  geom_col(position = position_stack(), colour = "white") +
  geom_text(aes(colour= fct_rev(freeway)), position = position_stack(vjust = .5))+
  scale_x_discrete(name = NULL)+
  guides(fill = guide_legend(reverse = TRUE)) +
  scale_y_continuous(name = NULL, expand = c(0,0),limits = c(0,131))+
  scale_fill_manual(name = NULL, values = c(rep(vago.grey,3), vago.blue))+
  scale_colour_manual(guide = "none", name = NULL, values = c("black","black","black","white"))+
  labs(title="Stacked column chart: 2 colours, single series highlight",
       subtitle = 
         str_wrap(
           "If you want to highlight one category and de-emphasise the remaining ones.",
           70)
       )+
  theme_vago_white()

# Add shares to data
car_accidents =
  car_accidents |> 
  group_by(month) |> 
  mutate(share = accidents/sum(accidents))

## example of single colour column plot
ggplot(data=car_accidents,aes(x=month, y = share, fill = fct_rev(freeway), label = scales::percent(share, accuracy = 1))) +
  geom_col(position = position_stack()) +
  geom_text(aes(colour= fct_rev(freeway)), position = position_stack(vjust = .5))+
  guides(fill = guide_legend(reverse = TRUE)) +
  scale_x_discrete(name = NULL)+
  scale_y_continuous(name = NULL, expand = c(0,0), label = scales::label_percent(), limits = c(0,1.01))+
  scale_fill_manual(name = NULL, values = rev(vago_colours$chart))+
  scale_colour_manual(guide = "none", name = NULL, values = c("black","black","black","white"))+
  labs(title="100% Stacked column chart: full colour",
       subtitle ="Effective to show relative contributions.")+
  theme_vago_white()

# Create data tibble to add forecaast variable
forecast = tibble(
  month = factor(c("Jan","Feb","Mar","Apr","May"), levels = month.abb),
  forecast = c("Actual","Actual","Actual","Forecast","Forecast"))

# Add to car_accidents using join
car_accidents =
  car_accidents |>
  left_join(forecast)
#> Joining with `by = join_by(month)`

## example of single colour column plot
ggplot(
  data=car_accidents,
  aes(
    x = month, 
    y = share, 
    alpha = forecast,
    fill = fct_rev(freeway),
    label = scales::percent(share, accuracy = 1))) +
  geom_col(position = position_stack()) +
  geom_text(
    aes(colour= fct_rev(freeway)),
    position = position_stack(vjust = .5))+
  guides(fill = guide_legend(reverse = TRUE)) +
  scale_x_discrete(name = NULL)+
  scale_alpha_manual(values = c(1,.6), guide = "none")+
  scale_y_continuous(name = NULL, expand = c(0,0), label = scales::label_percent())+
  scale_fill_manual(name = NULL, values = rev(vago_colours$chart))+
  scale_colour_manual(guide = "none", name = NULL, values = c("black","black","black","white"))+
  labs(title="100% Stacked column chart: full colour with forecast",
       subtitle ="You may consider using a shading pattern for forecast data.")+
  theme_vago_white()

Bar charts

Auditors use bar charts because they are a useful alternative to column charts, especially where category labels are lengthy, there’s no natural date sequence or there are only a few bars in the chart.

The order of the bars depends on the categories they depict: • when there is a natural sequence then typically column charts are preferred (for example, time series from left to right). • where there is no natural sequence, it may be appropriate to: - order categories alphabetically, especially if you wish to avoid perceptions of bias - order from highest to lowest (or lowest to highest), particularly if the issue is about demonstrating relativity between better practices and worst practices.

# Create data tibble
airport_flights = 
  tibble(
    city = c("London",  "Milan",    "Paris",    "Prague",   "Lisbon"),
    flights = c(45,16,  10, 10, 5)
  )

## example plot
ggplot(
  data = airport_flights, 
  aes(
    x = reorder(city, flights), 
    y = flights)) +
  geom_col(fill = vago.blue) +
  # Convert from columb to horizontal bar using coord_flip
  coord_flip()+
  geom_text(aes(label = flights), size = 5, nudge_y = 1)+
  scale_x_discrete(name = NULL)+
  scale_y_continuous(name = NULL, labels = NULL,  expand = c(0,0,.1,.1))+
  labs(title="Bar chart: one colour")+
  theme_vago_white()

mode_shares = 
  tibble(
    mode = 
      rep(
        c("Walk","Tram","Train","Bike","Car"),
        4) |> 
      factor(levels = c("Walk","Tram","Train","Bike","Car")),
    council = 
      c(rep("Council A",5),
        rep("Council B",5),
        rep("Council C",5),
        rep("Council D",5)),
    trips = 
      c(45, 16, 10, 10, 5, 
        40, 10, 12, 20, 10, 
        25, 18, 14, 30, 15, 
        15, 19, 18, 25, 20)
  )
# Create chart
ggplot(
  data = mode_shares, 
  aes(
    x = fct_rev(mode), 
    y = trips,
    fill = factor(council) |> fct_rev(),
    )
  )+
  geom_col() +
  geom_text(
    aes(label = trips,colour= factor(council) |> fct_rev()), 
    position = position_stack(vjust =.5))+
  # Convert from columb to horizontal bar using coord_flip
  coord_flip()+
  guides(fill = guide_legend(reverse = TRUE))+
  scale_x_discrete(name = NULL,  expand = c(0,0,.1,.1))+
  scale_y_continuous(
    name = NULL, expand = c(0,0,.1,.1), breaks = seq(0,140, by = 20))+
  scale_fill_manual(
    name = NULL, values = rev(vago_colours$chart))+
  scale_colour_manual(
    guide = "none", name = NULL, 
    values = c("black","black","white","white"))+
  labs(
    title="Stacked bar chart: full colour",
    subtitle = 
      str_wrap(
        "Use to show how each category contributes to the total and what percentage each represents.",
        70)
    )+
  theme_vago_white()

Line charts

Line charts display information as a series of data points connected by straight line segments. The measurement points are typically ordered by their x-axis value.

Line charts are best at showing how particular data changes at equal intervals of time. Use them to visualise a trend in data

Line charts emphasise time flow and rate of change, rather than the amount of change.

Key points: • Use markers to show your actual data points to distinguish the series from a continuous distribution. However, if there are many categories or the values are approximate, use a line chart without markers. • Use lines between markers to show trends. Use dashed lines if you want the focus to be on the markers. • Do not use a line chart if you are displaying categories on the X axis. Use a bar chart instead. • Do not use more than 2 colours in one chart. If the data series is a ‘nice to have’, push them to the background by using a dark grey colour.

# Line chart data 
line_data = 
  tibble(
    year = 
      rep(2020:2024,5),
    council = 
      c(rep("Council A",5),
        rep("Council B",5),
        rep("Council C",5),
        rep("Council D",5),
        rep("Target",5)),
    trips = 
      c(300,    150,    220,    260,    340,
        300,    190,    220,    400,    275,
        203,    157,    169,    223,    400,
        204,    187,    290,    325,  150,
        300,    300,    300,    300,    300)
  )

## example plot
ggplot(
  data = line_data |> filter(str_detect(council, "A")), 
  aes(
    x = year, 
    y = trips,
    fill = council,
    colour = council,
    )
  )+
  geom_point()+
  geom_line() +
  scale_x_continuous(name = NULL, breaks = 2020:2024, limits = c(2020,2024.4))+
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,410), breaks = seq(0,400, by = 100))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Single line",
       subtitle = "One-dimensional line: one colour, single axis")+
  theme_vago_white()

## example plot
ggplot(
  data = line_data |> filter(str_detect(council, "A|B")), 
  aes(
    x = year, 
    y = trips,
    fill = council,
    colour = council,
    )
  )+
  geom_point()+
  geom_line() +
  geom_text(
    data =  line_data |> filter(str_detect(council, "A|B"), year == max(year)),
    aes(label = council), nudge_y = 10) +
  scale_x_continuous(name = NULL, breaks = 2020:2024, limits = c(2020, 2024.4))+
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,510), breaks = seq(0,500, by = 100))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Two lines",
       subtitle = "Two-dimensional line: 2 colours, single axis")+
  theme_vago_white()

## example plot
ggplot(
  data = line_data |> filter(str_detect(council, "A")), 
  aes(
    x = year, 
    y = trips,
    fill = council,
    colour = council,
    )
  )+
  geom_line(
    data = line_data |> filter(str_detect(council, "Target")), 
    linetype = 2, 
    colour = vago.grey
    ) +
  geom_point()+
  geom_line() +
  geom_text(
    data =  line_data |> 
      filter(str_detect(council, "Target"), 
             year == max(year)),
    aes(label = council),
    colour = "black",
    size = 5,
    hjust = 0, 
    nudge_x = .05) +
  scale_x_continuous(name = NULL, breaks = 2020:2024, limits = c(2020,2024.4))+
  scale_y_continuous(
    name = NULL, expand = c(0,0), 
    limits = c(0,510), breaks = seq(0,500, by = 100))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Two lines",
       subtitle = "Two-dimensional line: 2 colours, single axis")+
  theme_vago_white()

## example plot
ggplot(
  data = line_data |> filter(str_detect(council, "A")), 
  aes(
    x = year, 
    y = trips,
    fill = council,
    colour = council,
    )
  )+
  geom_line(
    data = line_data |> filter(!str_detect(council, "A|Target")), 
    colour = vago.grey
  ) +
  geom_line() +
  geom_text(
    data =  line_data |> filter(str_detect(council, "A"), year == max(year)),
    aes(label = council), nudge_y = 10) +
  scale_x_continuous(name = NULL, breaks = 2020:2024, limits = c(2020,2020.4))+
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,510), breaks = seq(0,500, by = 100))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Multiple lines",
       subtitle = "One colour")+
  theme_vago_white()
#> Warning: Removed 12 rows containing missing values or values outside the scale range
#> (`geom_line()`).
#> `geom_line()`: Each group consists of only one observation.
#> ℹ Do you need to adjust the group aesthetic?
#> Warning: Removed 4 rows containing missing values or values outside the scale range
#> (`geom_line()`).
#> `geom_line()`: Each group consists of only one observation.
#> ℹ Do you need to adjust the group aesthetic?
#> Warning: Removed 1 row containing missing values or values outside the scale range
#> (`geom_text()`).

## example plot
ggplot(
  data = line_data |> filter(str_detect(council, "A|D")), 
  aes(
    x = year, 
    y = trips,
    fill = council,
    colour = council,
    )
  )+
  geom_line(
    data = line_data |> filter(!str_detect(council, "A|D|Target")), 
    colour = vago.grey
  ) +
  geom_line() +
  ggrepel::geom_text_repel(
    data =  line_data |> filter(str_detect(council, "A|D"), year == max(year)),
    aes(label = council), direction = "y", force = 40, min.segment.length = 20) +
  scale_x_continuous(name = NULL, breaks = 2020:2024, limits = c(2020,2024.4))+
  scale_y_continuous(name = NULL, expand = c(0,0), limits = c(0,510), breaks = seq(0,500, by = 100))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Multiple lines",
       subtitle = "Two colours")+
  theme_vago_white()

comparison_data =
  tibble(
    year = rep(c(2016:2024),2),
    type = c(rep("Expenditure",9),rep("Income",9)),
    value =
      c(42206844,   47883821,   60827924,   
        115072080, 138989959,   164836433,  
        191977494,  203121809,  231789110,
        23008170,   27059913,   32949213,   
        36212567,   40825086,   50466673,   
        60960692,   71814854,   100712099)
    )

## example plot
ggplot(
  data = comparison_data , 
  aes(
    x = year, 
    y = value/1e6,
    colour = type,
    )
  )+
  geom_line() +
  ggrepel::geom_text_repel(
    data =  comparison_data |> filter(year == max(year)),
    aes(label = type, colour = type), direction = "y", size = 5,
    force = 40, min.segment.length = 20) +
  scale_x_continuous(name = NULL, breaks = 2016:2024, limits = c(2016, 2024.4))+
  geom_segment(data = tibble(x = 2021.5, xend = 2021.5, y = 60, yend = 170, type = NA), 
               aes(x = x, y = y, xend = xend, yend=yend), colour = vago.grey, fill = NA)+
  annotate(geom = "text", size = 5, hjust = 0, label = "Difference = net cost", x = 2021.7, y = 115)+
  scale_y_continuous(
    name = "$ million", 
    expand = c(0,0), 
    limits = c(0,250), 
    breaks = seq(0,500, by = 50))+
  scale_fill_manual(name = NULL, guide = "none", values = (vago_colours$chart))+
  scale_colour_manual(name = NULL,guide = "none",  values = (vago_colours$chart))+
  labs(title="Two colours",
       subtitle = "Highlighting difference")+
  theme_vago_white()
#> Warning in geom_segment(data = tibble(x = 2021.5, xend = 2021.5, y = 60, :
#> Ignoring unknown parameters: `fill`

Maps

{vagotheme} exports a theme - theme_vago_map() to be used when plotting maps. This theme can be produced with white or lilax backgrounds. White is the default.

## basic example of a map with white background
ggplot(
  data = 
    strayr::read_absmap(
      name="sa42021",
      remove_year_suffix = TRUE
      ) |> 
    filter(state_name == "Victoria"),
  aes(
    geometry = geometry, 
    fill=state_name)
  ) +
  geom_sf() +
  scale_fill_manual(guide='none',values=vago.blue) +
  theme_vago_map(base_colour = "white", base_size = 10)

## basic example of a map with lilac background
ggplot(
  data = 
    strayr::read_absmap(
      name="sa42021",
      remove_year_suffix = TRUE
      ) |> 
    filter(state_name == "Victoria"),
  aes(
    geometry = geometry, 
    fill=state_name)
  ) +
  geom_sf() +
  scale_fill_manual(guide='none',values=vago.blue) +
  theme_vago_map(base_colour = "lilac", base_size = 10)
#> Reading sa42021 file found in C:\Users\SAMUEL~1\AppData\Local\Temp\Rtmpozi2ke

Documentation

See also documentation for the palettes package at https://mccarthy-m-g.github.io/palettes/ or in the installed package: help(package = "palettes").