Search code examples
rplotly

Treemaps With Plotly: Blank Screen

I am working with the R programming language.

I am trying to plot a Treemap with plotly, but it's returning a blank screen

I made my own dataset for this problem:

dados <- data.frame(
    Marca = c("Chevrolet", "Chevrolet", "Chevrolet", "Chevrolet", "Chevrolet", "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Ford", "Ford", "Ford", "Ford", "Ford", "Ford", "Honda", "Honda", "Hyundai", "Hyundai", "JAC", "Renault", "Renault", "Renault", "Renault", "Renault", "Volkswagen", "Volkswagen", "Volkswagen", "Volkswagen", "Volkswagen", "Volkswagen"),
    Carro = c("celta", "fox", "montana", "onix", "s10", "agile", "bravo", "cruze", "punto", "siena", "strada", "uno", "ecosport", "focus", "idea", "ka", "palio", "ranger", "city", "fit", "hb20", "tucson", "j3", "clio", "duster", "fluence", "logan", "sandero", "crossfox", "gol", "jetta", "polo", "saveiro", "voyage"),
    Quantidade = c(483, 609, 119, 144, 117, 389, 137, 147, 397, 371, 146, 185, 93, 217, 240, 195, 134, 34, 234, 134, 149, 304, 92, 300, 186, 242, 439, 608, 132, 618, 130, 217, 75, 315)
)

treemap <- plot_ly(
    data = dados,
    type = "treemap",
    labels = ~Carro,
    parents = ~Marca,
    values = ~Quantidade
)

treemap
Solution

  • You have to establish the parent labels; it's not inherent that it's the unique values of the data in Marca.

    Here's how you get Plotly to interpret this correctly, start by creating a summary of the data to get parent container totals and parent container labels. Then you'll combine your original data with this summary. Check it out:

    library(plotly)
library(tidyverse)

# summarize data frame to get parent container totals
d <- dados %>% group_by(Marca) %>% summarise(Quantidade = sum(Quantidade)) %>% 
  rename(Carro = Marca) %>%            # change columns
  mutate(Marca = "") %>%               # create blank column replacement
  select(names(dados))                 # put columns back in order

d2 <- rbind(d, dados) %>% as.data.frame() # combine with original data

    Now, when you call your plot, you'll call the data d2. You need one more argument, as well: branchvalues.

    treemap <- plot_ly(
  # data = dados,
  data = d2,
  type = "treemap",
  labels = ~Carro,
  parents = ~Marca,
  values = ~Quantidade,
  branchvalues = "total"
)

    enter image description here

    Here's all the code again, all together (easier copy + paste).

    library(plotly)
library(tidyverse)

dados <- data.frame(
  Marca = c("Chevrolet", "Chevrolet", "Chevrolet", "Chevrolet", "Chevrolet", 
            "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Fiat", "Ford",
            "Ford", "Ford", "Ford", "Ford", "Ford", "Honda", "Honda", "Hyundai", 
            "Hyundai", "JAC", "Renault", "Renault", "Renault", "Renault", 
            "Renault", "Volkswagen", "Volkswagen", "Volkswagen", "Volkswagen", 
            "Volkswagen", "Volkswagen"),
  Carro = c("celta", "fox", "montana", "onix", "s10", "agile", "bravo", "cruze",
            "punto", "siena", "strada", "uno", "ecosport", "focus", "idea", "ka", 
            "palio", "ranger", "city", "fit", "hb20", "tucson", "j3", "clio", 
            "duster", "fluence", "logan", "sandero", "crossfox", "gol", "jetta", 
            "polo", "saveiro", "voyage"),
  Quantidade = c(483, 609, 119, 144, 117, 389, 137, 147, 397, 371, 146, 185, 93,
                 217, 240, 195, 134, 34, 234, 134, 149, 304, 92, 300, 186, 242, 
                 439, 608, 132, 618, 130, 217, 75, 315)
)
# summarize data frame to get parent container totals
d <- dados %>% group_by(Marca) %>% summarise(Quantidade = sum(Quantidade)) %>% 
  rename(Carro = Marca) %>%            # change columns
  mutate(Marca = "") %>%               # create blank column replacement
  select(names(dados))                 # put columns back in order

d2 <- rbind(d, dados) %>% as.data.frame() # combine with original data

treemap <- plot_ly(
  # data = dados,
  data = d2, type = "treemap", labels = ~Carro,
  parents = ~Marca, values = ~Quantidade,
  branchvalues = "total"
)
treemap