The data this week comes from Kaggle and Kaggle courtesy of Georgios Karamanis.
The Mr Boston dataset was acquired from the Mr. Boston Bartender's Guide, while the cocktails.csv dataset was web-scraped as part of a hackathon.
These datasets are relatively clean, and have lots of interesting data to count, summarize, or possibly classify with a model!
While I have tamed both datasets, the Mr. Boston dataset is IMO cleaner, the cocktails.csv dataset was web-scraped and has some classic "funkiness" that you get from web-scraping (empty rows, now lines \n, etc). I've left both datasets for your exploration, where the cocktails.csv is probably better used for data cleaning/validation and the Mr. Boston appears to be cleaner and close to analysis-ready. However, there are additional columns in the cocktails.csv dataset, so maybe try joining or further cleaning! The cocktails.csv dataset also has more non-alcoholic drinks if you'd prefer to skip the alcohol.
- Margarita Clustering -- this doesn't use our dataset, but provides a potential idea for how to approach this data
- Information is Beautiful Graphic
- Recommender for Cocktail recipes
I've intentionally left the measure column as a string with a number + volume/unit so that you can try out potential strategies to cleaning it up.
Some potential tools:
The cocktails.csv dataset was web-scraped and has some classic "funkiness" that you get from web-scraping (empty rows, new lines \n with blank, etc)
# Get the Data
cocktails <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-26/cocktails.csv')
boston_cocktails <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-05-26/boston_cocktails.csv')
# Or read in with tidytuesdayR package (https://github.com/thebioengineer/tidytuesdayR)
# Either ISO-8601 date or year/week works!
# Install via devtools::install_github("thebioengineer/tidytuesdayR")
tuesdata <- tidytuesdayR::tt_load('2020-05-26')
tuesdata <- tidytuesdayR::tt_load(2020, week = 22)
cocktails <- tuesdata$cocktails
| variable | class | description |
|---|---|---|
| row_id | double | row identifier |
| drink | character | drink name |
| date_modified | double | date modified (web scraped) |
| id_drink | double | drink unique id |
| alcoholic | character | alcoholic, non alcoholic, optional |
| category | character | Category, eg cocktail, shot, etc |
| drink_thumb | character | thumbnail of the drink |
| glass | character | Recommended glass type |
| iba | character | International Bartenders association category |
| video | logical | Video to how to make |
| ingredient_number | integer | Ingredient number |
| ingredient | character | Ingredient |
| measure | character | Measurement/volume of ingredient |
| variable | class | description |
|---|---|---|
| name | character | Name of cocktail |
| category | character | Category of cocktail |
| row_id | integer | Drink identifier |
| ingredient_number | integer | Ingredient number |
| ingredient | character | Ingredient |
| measure | character | Measurement/volume of ingredient |
library(tidyverse)
# source for boston drinks
"https://www.kaggle.com/jenlooper/mr-boston-cocktail-dataset"
# Source for drinks
"https://www.kaggle.com/ai-first/cocktail-ingredients"
# Read in the data --------------------------------------------------------
drinks <- read_csv("2020/2020-05-26/all_drinks.csv") %>%
janitor::clean_names() %>%
rename(row_id = x1)
boston_drks <- read_csv("2020/2020-05-26/mr-boston-flattened.csv")
# pivot_longer drinks -----------------------------------------------------
drk_ing <- drinks %>%
select(row_id:str_iba, contains("ingredient"), str_video) %>%
# pivot to take wide data to long
pivot_longer(cols = contains("ingredient"),
names_to = "ingredient_number",
values_to = "ingredient") %>%
# remove text and extract only the digits
mutate(ingredient_number = str_extract(ingredient_number, "[:digit:]+") %>%
as.integer()) %>%
# remove "str_" from any of the col names
set_names(nm = str_remove(names(.), "str_"))
drk_measure <- drinks %>%
# select only the join ids and cols w/ "measure"
select(row_id, str_drink, id_drink, contains("measure")) %>%
# pivot to take wide data to long
pivot_longer(cols = contains("measure"),
names_to = "measure_number",
values_to = "measure") %>%
# extract just digits
mutate(measure_number = str_extract(measure_number, "[:digit:]+") %>%
as.integer()) %>%
# remove str_ from any col names
set_names(nm = str_remove(names(.), "str_"))
# join the two long dfs back together
all_drks <- left_join(drk_ing, drk_measure,
by = c("row_id", "drink", "id_drink",
"ingredient_number" = "measure_number")) %>%
filter(!is.na(measure) & !is.na(ingredient))
# confirm if missing data
# confirm if missing data
anti_join(drk_ing, drk_measure,
by = c("row_id", "drink", "id_drink",
"ingredient_number" = "measure_number"))
write_csv(all_drks, "2020/2020-05-26/cocktails.csv")
# pivot_longer boston drinks ----------------------------------------------
bs_drk_ing <- boston_drks %>%
mutate(row_id = row_number()) %>%
select(name, category, row_id, contains("ingredient")) %>%
pivot_longer(cols = contains("ingredient"),
names_to = "ingredient_number",
values_to = "ingredient") %>%
mutate(ingredient_number = str_extract(ingredient_number, "[:digit:]+") %>%
as.integer())
bs_drk_ms <- boston_drks %>%
mutate(row_id = row_number()) %>%
select(name, category, row_id, contains("measurement")) %>%
pivot_longer(cols = contains("measurement"),
names_to = "measure_number",
values_to = "measure") %>%
mutate(measure_number = str_extract(measure_number, "[:digit:]+") %>%
as.integer())
all_bs_drks <- left_join(bs_drk_ing, bs_drk_ms,
by = c("name", "category", "row_id",
"ingredient_number" = "measure_number")) %>%
filter(!is.na(ingredient) & !is.na(measure))
# confirm if missing data
anti_join(bs_drk_ing, bs_drk_ms,
by = c("name", "category", "row_id",
"ingredient_number" = "measure_number"))
write_csv(all_bs_drks, "2020/2020-05-26/boston_cocktails.csv")