Recoding and Joining

Author

Josef Fruehwald

Published

February 8, 2023

More working with babynames

Load up the libraries

Read in a dataframe of automatically generated pronunciations.

name_pronunciation <- read_csv("https://jofrhwld.github.io/AandS500_2023/data/name_pronunciation.csv")

Recapping

We need to do some data processing to get it into a format that works well with plotting.

babynames |>
  filter(name == "Mary") |>
  ggplot(aes(x = year, y= n, color = sex)) +
    geom_line() +
    scale_color_brewer(palette = "Dark2")

Adding new columns and summarizing

babynames |>
  mutate(
    fake_col = "a",
    ends_in_n = str_ends(name, "n")
  ) |>
  summarise(
    .by = c(year, sex, ends_in_n),
    total = sum(n)
  ) |>
  ggplot(aes(year, total, color = ends_in_n))+
    geom_line() +
    facet_wrap(~sex)

Re-coding

The function case_when() will recode data. What you give it is

<logical statement> ~ <new value>

It evaluates the logical statements in sequence.

data.frame(
  name = c("Joe", "Paul", "Kate", "Rebecca", "Lil")
) |>
  mutate(
    ends_in = case_when(
      name == "Paul" ~ "l",
      str_ends(name, "e") ~ "e",
      .default = "other"
    )
  )
     name ends_in
1     Joe       e
2    Paul       l
3    Kate       e
4 Rebecca   other
5     Lil   other

Recoding starwars characters’ heights into tall, medium, short.

starwars |>
  select(name, height) |>
  mutate(
    height_category = case_when(
      height >= 200 ~ "tall",
      height >= 100 ~ "medium",
      height >= 0 ~ "short",
      .default = "unrecorded"
    )
  ) |>
  arrange(desc(height)) |>
  summarise(
    .by = height_category,
    average = mean(height, na.rm = T),
    n = n()
  )
# A tibble: 4 × 3
  height_category average     n
  <chr>             <dbl> <int>
1 tall               220.    11
2 medium             176.    63
3 short               88      7
4 unrecorded         NaN      6

whose height isn’t there

starwars |>
  filter(is.na(height))
# A tibble: 6 × 14
  name         height  mass hair_…¹ skin_…² eye_c…³ birth…⁴ sex   gender homew…⁵
  <chr>         <int> <dbl> <chr>   <chr>   <chr>     <dbl> <chr> <chr>  <chr>  
1 Arvel Crynyd     NA    NA brown   fair    brown        NA male  mascu… <NA>   
2 Finn             NA    NA black   dark    dark         NA male  mascu… <NA>   
3 Rey              NA    NA brown   light   hazel        NA fema… femin… <NA>   
4 Poe Dameron      NA    NA brown   light   brown        NA male  mascu… <NA>   
5 BB8              NA    NA none    none    black        NA none  mascu… <NA>   
6 Captain Pha…     NA    NA unknown unknown unknown      NA <NA>  <NA>   <NA>   
# … with 4 more variables: species <chr>, films <list>, vehicles <list>,
#   starships <list>, and abbreviated variable names ¹​hair_color, ²​skin_color,
#   ³​eye_color, ⁴​birth_year, ⁵​homeworld

Joining

Joining together datasets that have a shared “key”.

name_pronunciation <- read_csv("https://jofrhwld.github.io/AandS500_2023/data/name_pronunciation.csv")

This is a dataframe with a name column shared withbabynames and pronunciation guesses.

name_pronunciation |> head()
# A tibble: 6 × 2
  name      name_pronounce          
  <chr>     <chr>                   
1 Mary      M EH1 R IY0             
2 Anna      AE1 N AH0               
3 Emma      EH1 M AH0               
4 Elizabeth IH0 L IH1 Z AH0 B AH0 TH
5 Minnie    M IH1 N IY0             
6 Margaret  M AA1 R G ER0 IH0 T     
babynames |> head()
# A tibble: 6 × 5
   year sex   name          n   prop
  <dbl> <chr> <chr>     <int>  <dbl>
1  1880 F     Mary       7065 0.0724
2  1880 F     Anna       2604 0.0267
3  1880 F     Emma       2003 0.0205
4  1880 F     Elizabeth  1939 0.0199
5  1880 F     Minnie     1746 0.0179
6  1880 F     Margaret   1578 0.0162

Using left_join() will return every row from the “left hand” data frame, and ever matching value from the “right hand” data frame.

babynames |>
  left_join(name_pronunciation)->
  bn_with_pron

bn_with_pron |> head()
# A tibble: 6 × 6
   year sex   name          n   prop name_pronounce          
  <dbl> <chr> <chr>     <int>  <dbl> <chr>                   
1  1880 F     Mary       7065 0.0724 M EH1 R IY0             
2  1880 F     Anna       2604 0.0267 AE1 N AH0               
3  1880 F     Emma       2003 0.0205 EH1 M AH0               
4  1880 F     Elizabeth  1939 0.0199 IH0 L IH1 Z AH0 B AH0 TH
5  1880 F     Minnie     1746 0.0179 M IH1 N IY0             
6  1880 F     Margaret   1578 0.0162 M AA1 R G ER0 IH0 T     
bn_with_pron |>
  filter(name == "Lee") |> 
  head()
# A tibble: 6 × 6
   year sex   name      n     prop name_pronounce
  <dbl> <chr> <chr> <int>    <dbl> <chr>         
1  1880 F     Lee      28 0.000287 L IY1         
2  1880 M     Lee     361 0.00305  L IY1         
3  1881 F     Lee      39 0.000395 L IY1         
4  1881 M     Lee     342 0.00316  L IY1         
5  1882 F     Lee      43 0.000372 L IY1         
6  1882 M     Lee     427 0.00350  L IY1

Now we can filter by pronunciation and look at the most popular spellings

bn_with_pron |>
  filter(name_pronounce == "L IY1") |>
  summarise(
    .by = c(year, sex, name),
    total = sum(n)
  ) |>
  ggplot(aes(year, total, color = name))+
    geom_line()+
    facet_wrap(~sex)

We didn’t get to this.

# name_pronunciation |>
#   select(name_pronounce) |>
#   distinct() |>
#   mutate(nsyl = str_count(name_pronounce, r"([AEIOU].\d)"))