ggplot_till.Rmd

``` r
library(ggplot2)  # data vis
library(dplyr)    # data manipulation
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union
library(readr)    # data io

# set a theme for the plot
theme_set(theme_light())

df <- read_csv("~/Downloads/dmwe.csv")
#> Parsed with column specification:
#> cols(
#>   countrycode = col_character(),
#>   SE_HS = col_character(),
#>   gdppercap = col_character(),
#>   general = col_character()
#> )

# some basic cleaning
df <- df %>% 
  mutate(
    SE_HS     = as.numeric(SE_HS),
    gdppercap = as.numeric(gdppercap),
    general   = as.numeric(general),
    # compute the cut before the ggplot! easier debugging
    cut_gdp   = cut(gdppercap, c(-Inf, 15000, Inf)),
    # a second, faster/easier approach: compute T/F if the value is smaller than 15k
    cut_gdp2  = gdppercap <= 15000
  )

# first plot
ggplot(df, aes(x = general, y = SE_HS, color = cut_gdp)) +
  geom_point(size = 3) +
  scale_color_manual(name = "gdppercap",
                     # notice the 1.5e+04 instead of 15000! you adressed the wrong factor-labels
                     values = c("(-Inf,1.5e+04]" = "black",
                                "(1.5e+04, Inf]" = "red"),
                     labels = c("<= 15000", "> 15000"))
#> Warning: Removed 14105 rows containing missing values (geom_point).
```

![](https://i.imgur.com/ltvxGEK.png)

``` r

# a second plot without the missing values
df %>% 
  filter(!is.na(cut_gdp2)) %>% 
ggplot(aes(x = general, y = SE_HS, color = cut_gdp2)) +
  geom_point(size = 3) +
  scale_color_manual(name = "gdppercap",
                     values = c("TRUE" = "black",
                                "FALSE" = "red"),
                     labels = c("TRUE" = "<= 15000", "FALSE" = "> 15000"))
#> Warning: Removed 8725 rows containing missing values (geom_point).
```

![](https://i.imgur.com/QaGCSgS.png)