Pizzanalysis
Insights from scraped menus
We scraped Just Eat
Why?
I needed a project for a .pizza domain
How?
POST /android.php?
var0=effb75066e9800f8960cce6a4c13f947&
var1=getrestaurants&
…
var4=55.6760968&
var5=12.5683371
…
var0 = MD5(SECRET+LAT+LNG+...)
SECRET=4ndr01d
272,498 products
name | description | price | category | restaurant_id |
Spice pizza [Alm.] | Med kebab, salat og creme fraiche dressing | 83.00 | Pizzaer | ON33N5PN |
Let’s explore the data!
“with kebab, salad and creme fraiche dressing“
[
“kebab”,
“salad”, “creme_fraiche_dressing”]
pizza2vec
“Ingredients used together are closer together in space”
pineapple
pesto
parmesan
What is the closest ingredient to pineapple?
peas
Which ingredients are most similar to pesto?
mascarpone, mozzarella, olive oil
mascarpone, mozzarella, olive oil
(fancy pizzas)
Ingredient clusters
What value does an ingredient add?
“Lets predict prices using ingredients”
Margarita index
Margerita index
Margharita index
Margherita index
Margh?[ae]rita index
Margherita index
Further research
Magenta special:
Hvad tror I en stavefejl koster?
Hvor meget billigere er en
“Pizza med killing”
end en
“Pizza med kylling”?
Ca. 2 kr. pr. stavefejl
github.com/volesen/just-scrape