geom_bar vs geom_col

geom_bar_col.Rmd

---
title: "geom_col vs geom_bar"
author: "Martin Monkman"
date: "2020/04/19"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
```


Gina Reynolds tweeted this:

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Here, <a href="https://t.co/px20P5hDI7">https://t.co/px20P5hDI7</a>, you can get a sense of how geom_bar and geom_col are different, and what&#39;s geom_area. I&#39;ve been wanting to do some work on &quot;stat&quot;... NHL (hockey) actually presents some interesting challenge/question because of year canceled! <a href="https://t.co/kkJyI2aSbb">pic.twitter.com/kkJyI2aSbb</a></p>&mdash; Gina Reynolds (@EvaMaeRey) <a href="https://twitter.com/EvaMaeRey/status/1251586868841336832?ref_src=twsrc%5Etfw">April 18, 2020</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> 

And then replied with this:

<blockquote class="twitter-tweet"><p lang="en" dir="ltr"><a href="https://twitter.com/andrewheiss?ref_src=twsrc%5Etfw">@andrewheiss</a> <a href="https://twitter.com/CedScherer?ref_src=twsrc%5Etfw">@CedScherer</a> can you articulate why aes(fill = season) doesn&#39;t &quot;do&quot; anything in the geom_bar example (follows if you click through above link). I&#39;m having trouble putting my finger on it.</p>&mdash; Gina Reynolds (@EvaMaeRey) <a href="https://twitter.com/EvaMaeRey/status/1251587737385230336?ref_src=twsrc%5Etfw">April 18, 2020</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> 


I thought I'd play around a bit with some of the other geoms and variables to try to understand what's happening.


## Original code


```{r data_read}
readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-03-03/game_goals.csv') ->  
  game_goals  

```


### Gina's original plots

```{r plot_col}

game_goals %>%  
  count(season) %>%  
  full_join(tibble(season = 1980:2020)) %>%  
  ggplot() +  
  aes(x = season) +  
  aes(y = n) +  
  geom_col() +  # stat is identity
  geom_point() +  # stat is identity
  geom_line() +  # stat is identity
  geom_area(alpha = .2,  # stat is identity
            fill = "magenta") +  
  aes(fill = season) +  
  aes(xend = season) +  
  aes(yend = 0) +  
  geom_segment(linetype = "dotted") 

#->  
#you_aggregate_then_plot  

```


```{r plot_bar}

game_goals %>%  
  ggplot() +  
  aes(x = season) +  
  geom_bar() +  # a version of geom column
  geom_point(stat = "count") +  
  geom_line(stat = "count") +  
  geom_area(stat = "count",  
            alpha = .2,  
            fill = "magenta") +  
  aes(fill = season)   # not completely grasping why fill cant work here

```


## tests

(Note: simplified code with just geom_col and geom_bar)

```{r test_col}
game_goals %>%  
  count(season) %>%  
  full_join(tibble(season = 1980:2020)) %>%  
  ggplot() +  
  aes(x = season) +  
  aes(y = n) +  # need to specify a y-axis variable
  geom_col() +
  aes(fill = season)  

```

```{r test_bar}
game_goals %>%  
  ggplot() +  
  aes(x = season) + 
  geom_bar()


```


### Try other variables

```{r test_bar}
game_goals %>%  
  ggplot() +  
  aes(x = season, fill = team) +  # add `fill = team`
  geom_bar()

game_goals %>%  
  ggplot() +  
  aes(x = season) +  
  geom_bar() +
  aes(fill = team)  # add `fill = team`


```


```{r test_smooth}
# geom_col -- works
game_goals %>%  
  count(season) %>%  
  full_join(tibble(season = 1980:2020)) %>%  
  ggplot() +  
  aes(x = season) +  
  aes(y = n) +  # need to specify a y-axis variable
  geom_col() +
  geom_smooth()


# geom_bar -- fails
game_goals %>%  
  ggplot() +  
  aes(x = season) +  
  geom_bar() +
  geom_smooth() # "Error: stat_smooth requires the following missing aesthetics: y"


```