X-axis max doubled when adding stacked e_bars?

[sharlagelfand]

Hi! Thanks for this great package :)

I'm running into a strange issue where the x-axis max is doubled (?) when I'm adding two e_bar()s that are stacked. I want to create a bar chart where one bar is "highlighted", so I'm accomplishing that with two series.

The x limits in the data are the same, and they each appear separately just fine:

library(echarts4r)
library(dplyr)

df_1 <- data.frame(
  x = 1:5, 
  y_1 = c(1, 2, 3, NA, 5)
)

df_1 %>%
  e_chart(x) %>%
  e_bar(y_1)

df_2 <- data.frame(
  x = 1:5, 
  y_2 = c(NA, NA, NA, 4, NA)
)

df_2 %>%
  e_chart(x) %>%
  e_bar(y_2)

When I combine, the styling looks as desired - I'm "fake" stacking them I guess, because I don't want the dodged behaviour, and that looks fine - but the x-axis max is doubled, going all the way to 10 instead of 5:

df_combined <- df_1 %>%
  left_join(df_2, by = "x")

df_combined %>%
  e_chart(x) %>%
  e_bar(y_1, stack = "group") %>%
  e_bar(y_2, stack = "group")

Any ideas why this might be? I'm using the latest version from github. I know you can manually set the limits, but it's hard to do that on a dynamic / not-manual basis. Thank you!

[JohnCoene]

If that's of any consolation, it was difficult for me to figure out too.

In Echarts.js the bar serie is "optimised" for a categorical x axis, it never really works right with a number type on the x axis.
Internally, echarts4r infers the type it sets in Echarts.js based on the data it is being fed.
Converting x to factor or categorical corrects this.

df_combined %>%
  mutate(x = as.factor(x)) %>% 
  e_chart(x) %>%
  e_bar(y_1, stack = "group") %>%
  e_bar(y_2, stack = "group")

With the above echarts4r observes that the data on the x axis is factor and correctly communicates to Echarts.js that
the x axis should be categorical.

Let me know if this is not clear.
I should probably makes this clearer in the documentation.

[JohnCoene]

I've added a note about it in the documentation.

I should perhaps force the data type to categorical internally but I'm worried this may impact other things. I'll think about it.

Thanks again for reporting this.

[sharlagelfand]

Ahhh that makes sense, thanks for clarifying! Looks great now.

Also helps me understand why manually setting the limits didn't really do what I want - was definitely acting factor-like! Thanks!