Stacked Bar Chart With No Fill?

Asked by At

I have a dataset all set up and am just trying to get a stacked bar chart to show.

My X-Axis will show three bars: "left", "middle", and "right"

My Y-Axis will be the "total_completed_epa" associated with each of the bars.

The only problem is that "total_completed_epa" is a mutated variable created by the sum of two other columns in my data-set. I just want to show the proportion that each of the columns had on the "total_completed_epa" with a stacked bar chart.

The data is:


pass_location Air_Epa_Play YAC_EPA_Play Total_Completed_EPA
middle         0.263         0.434           0.697
left           0.086         0.439           0.525
right          0.082         0.442           0.524

Scraping The Data:

pass_epa <- pbp_2018 %>%
  filter(play_type %in% c("pass", "no_play", "qb_spike"),
         epa != is.na(epa)) %>%
  group_by (pass_location) %>%
  summarize(pass_epa = sum(epa),
            air_epa = sum(comp_air_epa),
            yac_epa = sum(comp_yac_epa),
            pass_plays = n ()) %>%
  ungroup() %>% 
  mutate(EPA_Play = round(pass_epa / pass_plays, 3),
         Air_Epa_Play = round(air_epa / pass_plays, 3), 
         YAC_EPA_Play = round(yac_epa / pass_plays, 3),
         Total_Completed_EPA = Air_Epa_Play + YAC_EPA_Play) %>%
  slice(-1) %>% 
  arrange(-EPA_Play) %>% 
  filter(pass_plays >= 80) %>%
  select(pass_location, Air_Epa_Play, YAC_EPA_Play, Total_Completed_EPA) 

So I have just the 4 columns in my data set. The Air_Epa_Play and YAC_EPA_Play is summed into a "Total_Completed_EPA".

Visualizing It:

ggplot(pass_epa, aes(x = pass_location, y = Total_Completed_EPA, fill = ?)) +
  geom_col(position = "dodge") 

I just cannot get the Air_EPA_Play and YAC_EPA_Play to stack in the barchart for Total_Completed_EPA

enter image description here

1 Answers

1
CPak On

Data

test <- data.frame(pass_location=c('middle','left','right'), Air_Epa_Play=c(0.263,0.086,0.082), YAC_Epa_Play=c(0.434,0.439,0.442), Total_Completed=c(0.697,0.525,0.524))

  pass_location Air_Epa_Play YAC_Epa_Play Total_Completed
1        middle        0.263        0.434           0.697
2          left        0.086        0.439           0.525
3         right        0.082        0.442           0.524

You can ignore the Total_Completed column - select(-Total_Completed). ggplot does the stacking/summation for you so you don't have to calculate the total yourself. However, ggplot also likes data in long format (rather than wide format), so you'll need to gather() the relevant values (on y-axis) into a single column. Notice I use gather(..., -pass_location) to ignore the grouping column. Try the following with and without fill=var. Once you see that ggplot likes long formatted data, it becomes more intuitive to use it - at least it did for me.

library(tidyverse)
test %>% 
  select(-Total_Completed) %>% 
  gather(var, value, -pass_location) %>% 
  ggplot(., aes(x=pass_location, y=value, fill=var)) + 
  geom_col()