[OC] Part 2: Timing of When Ingredients are Added to Dishes + More Ingredients, Splits by Cuisine and Dish Type, and an Interactive Tool

[Part 1](https://www.reddit.com/r/dataisbeautiful/comments/1v9bq7i/oc\_timing\_of\_when\_ingredients\_are\_added\_to\_recipes/)

This is a follow up to an earlier post — description will be below. I created visualizations to explain garlic before onion, lemon showing up everywhere, and how certain ingredients differ wildly by cuisine type and dish type. I also built a tool where you can inspect any ingredient you want given there is sufficient support.

i was still surprised by lemon showing up pretty constantly in main dishes; I discovered it was mostly marinades and dressings.

let me know anything that still seems strange in this dataset or any improvements I can make!

Dataset and Tools:

250 cleaned canonical recipe dataset : https://palate.kitchen/data

Interactive Tool: https://palate.kitchen/instrument/when-to-add

recipe1m+ and python + an LLM sweep to prune garbage recipes from the dataset I pruned 20,000 recipes down to about 3000 'canonical recipes.' Regex + massive KV mapping (e.g. beef, ground beef, 80/20 beef are all the same) for ingredient normalization.

original post and context:

I scraped a few thousand canonical recipes across 45 cuisines then analyzed how early (or late) each ingredient showed up in recipes steps.

The photos shows a small but representative sample of the ingredients analyzed

There's a lot of science behind when you should add ingredients to recipes. This involves fat-solubility and the delicacy of some volatile compounds. e.g. cardamon seeds are fat-soluble so they should be added early while basil has very delicate aroma compounds so it should be added late.

But I honestly find that hard to intuitively learn for new ingredients, so this is a rough estimate and interesting viz to get a quick understanding of when to add an ingredient