Intro

Heading 2

wb_income_groups %>% head()

Cool Charts

Here’s a basic chart

emissions_dataset %>%
  # filter for latest data
  filter(year == 2019) %>%
  # get the top 5 by cumulative Co2 per capita 
  slice_max(order_by = cumulative_co2_per_capita, n = 5) %>%
  # fct_reorder will sort x by cumulative Co2 per capita
  ggplot(aes(x = fct_reorder(.f = country_name,.x = cumulative_co2_per_capita, 
                             .desc = TRUE), 
              # fill will change the color by whether a country is EM or DM
             y = cumulative_co2_per_capita, fill = em_dm)) +
  # just get used to the `stat = identity` thing for bar charts, default is `count`
  geom_bar(stat = "identity")

This is good enough if we’re just playing around with the data for our own understanding. But you wouldn’t want to share this with others. It’s ugly. We don’t know what we’re looking at. We don’t know the units. We don’t know where the data comes from.

But cleaning it up is pretty easy.

Here’s that chart cleaned up for presentation

emissions_dataset %>%
  # filter for latest data
  filter(year == 2019) %>%
  # get the top 5 by cumulative Co2 per capita 
  slice_max(order_by = cumulative_co2_per_capita, n = 5) %>%
  # fct_reorder will sort x by cumulative Co2 per capita
  ggplot(aes(x = fct_reorder(.f = country_name,.x = cumulative_co2_per_capita, 
                             .desc = TRUE), 
              # fill will change the color by whether a country is EM or DM
             y = cumulative_co2_per_capita, fill = em_dm)) +
  # just get used to the `stat = identity` thing for bar charts, default is `count`
  geom_bar(stat = "identity") +
  geom_text(aes(label = format(cumulative_co2_per_capita, digits = 0, big.mark = ",")), vjust = -0.2) + 
  # expand the y limits so the text doesn't get cut off
  scale_y_continuous(limits = c(0, 1500), breaks = scales::pretty_breaks(n = 10)) +
  # you can use color names or hex codes
  scale_fill_manual(values = c("Developed Markets" = "grey", "Emerging Markets" = "#008080")) +
  # a minimal theme with mostly white space
  theme_minimal() +
  labs(title = "Wow, this is profound",
       subtitle = "and some more context",
       x = "", # it's obvious this is country
       y = "Tons Co2 per capita",
       fill = "", # this gets rid of the `em_dm` tag on top of the legend,
       caption = "Data from x, Analysis by us"
       ) +
  # moves the legend from the side (default) to the bottom.
  theme(legend.position = "bottom")

Now we’ve cleaned it up into a format that we can share. While this may look like a lot of code, remember that good programmers work hard to be lazy.

  • Most of the tricks for making your chart look pretty and professional are shown in this code in some form. It’s not point and click easy, but it’s not too much to learn with a little bit of practice.
  • If you use this code often, you can turn it into simple charting functions that will automatically make your charts look spiffy and professional with one line of code.

More Resources

Here are some great resources for making great ggplot2 charts