This is a further exploration of the halloween candy dataset taken from the fivethirtyeight article "The Ultimate Halloween Candy Power Ranking"
link to The Ulitmate Halloween Candy Power Ranking
In this article they explored the desirabiltiy of 85 common Halloween treats
They focused on properties of the treats to see if they could identify what made them desirable
- Chocolate
- Fruity
- Caramel
- Nuts
- Nougat
- Crispy
- Hard
#Read in the data
file_path <- "https://raw.githubusercontent.com/catfoodlover/Data607/main/Data607_HW1_Data_WilliamAiken.csv"
candy_df <- read_csv(file_path, show_col_types = FALSE)
#Select columns of interest and fix names when necessary
candy_df <- candy_df %>% select(names = competitorname, choco = chocolate, fruity, caramel, nuts = peanutyalmondy, nougat, crispy = crispedricewafer, win_percent = winpercent)
#Character needs to be fixed
candy_df$names <- gsub( "Õ", "'", candy_df$names)
#Convert dataset from wide to long to make it easier to work with
temp <- melt(data = candy_df, id.vars = "names", measure.vars = c("choco", "fruity", "caramel", "nuts", "nougat", "crispy"), variable.name = "Property", value.name = "Status")
#Join win percentage back in
temp2 <- left_join(temp, candy_df %>% select(names, win_percent), by = "names")
What is the mean win percentage of all properties?
- People like texture, nuts and crispy are stand outs both those are both properties that tend to go with chocolate the 3rd place property.
temp2 %>% filter(Status == 1) %>% select(Property, win_percent) %>% tbl_summary(by = Property,
statistic = list(all_continuous() ~ "{mean} ({sd})"),
digits = all_continuous() ~ 1,
label = win_percent ~ "% Winner")
| Characteristic |
choco, N = 37 |
fruity, N = 38 |
caramel, N = 14 |
nuts, N = 14 |
nougat, N = 7 |
crispy, N = 7 |
| % Winner |
60.9 (12.8) |
44.1 (10.3) |
57.3 (16.2) |
63.7 (16.4) |
60.1 (13.8) |
66.2 (10.7) |
what are the properties of the top bottom candies?
- Looks like people love chocolate.
candy_df %>% arrange(desc(win_percent)) %>% slice_head(n = 10) %>%
count(choco, fruity, caramel, nuts, nougat, crispy) %>% kable() %>% kable_styling()
|
choco
|
fruity
|
caramel
|
nuts
|
nougat
|
crispy
|
n
|
|
1
|
0
|
0
|
0
|
0
|
1
|
1
|
|
1
|
0
|
0
|
1
|
0
|
0
|
6
|
|
1
|
0
|
1
|
0
|
0
|
1
|
1
|
|
1
|
0
|
1
|
0
|
1
|
0
|
1
|
|
1
|
0
|
1
|
1
|
1
|
0
|
1
|
What are the properties of the bottom ten treats?
- There a less clear message. Fruity isn't popular but more importantly none of these are chocolate.
candy_df %>% arrange(desc(win_percent)) %>% slice_tail(n = 10) %>%
count(choco, fruity, caramel, nuts, nougat, crispy) %>% kable() %>% kable_styling()
|
choco
|
fruity
|
caramel
|
nuts
|
nougat
|
crispy
|
n
|
|
0
|
0
|
0
|
0
|
0
|
0
|
3
|
|
0
|
0
|
0
|
1
|
0
|
0
|
1
|
|
0
|
0
|
1
|
0
|
0
|
0
|
2
|
|
0
|
1
|
0
|
0
|
0
|
0
|
4
|