Europe Energy

Using echarts4r to create a treemap of nested types of energy usage.

Tidy Tuesday: 4 August 2020

This week’s data can be found here. It is about energy usage in Europe.

energy_types <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-08-04/energy_types.csv')

energy_types %>%
  dplyr::group_by(country_name) %>%
  dplyr::summarise_if(is.numeric, sum) %>%
  echarts4r::e_charts(country_name) %>% 
  echarts4r::e_parallel(`2016`, `2017`, `2018`) %>% 
  echarts4r::e_theme('wonderland') %>%
  echarts4r::e_title("Total energy usage by country")

I would like to visualise this data as a treemap using the echarts4r package. Treemaps visualise parts of a whole, and can be grouped. For example, in this dataset there are two levels of hydropower: there’s a row for total hydropower, and a row for pumped hydropower which is a sub-set of the total. Treemaps can show such nested parts-of-a-whole simultaneously. Documentation on the package can be found here.

Let’s separate hydro power into correctly named rows: pumpedhydro power, and other hydropower.

hydro_prep <- energy_types %>%
  dplyr::filter(grepl('hydro', tolower(type))) %>%
  tidyr::pivot_wider(id_cols = country:country_name,
                     names_from = type,
                     values_from = `2016`:`2018`) %>%
  dplyr::mutate(`2016_Other hydro` = `2016_Hydro` - `2016_Pumped hydro power`,
                `2017_Other hydro` = `2017_Hydro` - `2017_Pumped hydro power`,
                `2018_Other hydro` = `2018_Hydro` - `2018_Pumped hydro power`) %>%
  tidyr::pivot_longer(cols = `2016_Hydro`:`2018_Other hydro`,
                      names_to = c('year', 'type'),
                      names_sep = '_',
                      values_to = 'power') %>%
  dplyr::filter(type != 'Hydro') %>%
  dplyr::transmute(parent = 'Hydro',
                child = type,
                value = power,
                year,
                country_name)
head(hydro_prep)

# A tibble: 6 x 5
  parent child              value year  country_name
  <chr>  <chr>              <dbl> <chr> <chr>       
1 Hydro  Pumped hydro power 1110  2016  Belgium     
2 Hydro  Pumped hydro power 1093. 2017  Belgium     
3 Hydro  Pumped hydro power  983. 2018  Belgium     
4 Hydro  Other hydro         366  2016  Belgium     
5 Hydro  Other hydro         268. 2017  Belgium     
6 Hydro  Other hydro         256. 2018  Belgium     

The other energy types don’t have sub-types; thus we can simply transmute its columns into the right shape.

other_prep <- energy_types %>%
  dplyr::filter(!grepl('hydro', tolower(type))) %>%
  tidyr::pivot_longer(cols = `2016`:`2018`,
                      names_to = 'year',
                      values_to = 'power') %>%
  dplyr::transmute(parent = type,
                child = type,
                value = power,
                year,
                country_name)

To see the best impact of the treemap visualisation, let’s visualise the country that uses a mix of energy types, but with significant hydropower. We can select this based ont he 2018 energy usage with some filtering and arranging.

most_hydro <- energy_types %>%
  dplyr::filter(level == 'Level 1') %>%
  tidyr::pivot_wider(id_cols = country_name,
                     names_from = type,
                     values_from = `2018`) %>%
  dplyr::arrange(-Hydro) %>%
  dplyr::filter(Nuclear != 0, Geothermal != 0) %>%
  slice(1) %>%
  .$country_name

To create year-by-year treemaps, we can use a loop to filter the dataframe and create separate treemaps for separate years. The e_arrange function then puts them all in a row for us.

years <- sort(unique(hydro_prep$year))
data <- rbind(other_prep, hydro_prep) %>%
  dplyr::filter(country_name == most_hydro)
for(year_num in years){
  data_filter <- data %>%
    dplyr::filter(year == year_num) 
  e <- data_filter %>%
    echarts4r::e_charts() %>%
    echarts4r::e_treemap(parent, child, value) %>% 
    echarts4r::e_title(paste0("Energy in ", most_hydro, ", ", year_num)) %>%
    echarts4r::e_theme('wonderland')
  assign(paste0('treemap_', year_num), e)
}
echarts4r::e_arrange(treemap_2016, treemap_2017, treemap_2018, cols = 3, rows = 1)