3 Million Instacart Orders, Open Sourced

Just getting started with the lessons, so I can’t really point out any specific application regarding the course, but I thought this might be interesting to play with:


Very cool dataset. Also interested in one of the libraries hey mention they use in production:

1 Like

this is fantastic! Thanks for sharing the dataset and the annoy library.

@stefanwild, @davecg Are you aware of an instacart competition? I ask because it showed up in their data dictionary.

  • “train”: training data supplied to participants (~131k orders)
  • “test”: test data reserved for machine learning competitions (~75k orders)

This Instacart dataset is cool. Anyone here has played with basket of goods analysis either professionally or for fun?

I am looking for the ‘real-world’ questions that are worth asking if you worked at asking if you worked at a retailer. Especially, the tough ones that cannot be solved via traditional means.

Their previous blog on how to speed up order pickup is also a good read.