Choose data sets used in app examples

[rpodcast]

Ideally we can find data sets that have broad appeal and easily accessible.

[rpodcast]

Assorted ideas after some long searching:

• The Instacart Online Grocery Shopping Dataset 2017 https://tech.instacart.com/3-million-instacart-orders-open-sourced-d40d29ead6f2
• Board games dataset https://www.kaggle.com/datasets/gabrio/board-games-dataset
• Public benefit corporation data https://data.world/blab/b-corp-impact-data/workspace/data-dictionary

[rpodcast]

Thanks to @chendaniely , we have a new contender: The Metropolitan Museum of Art Collection API and associated metadata. Important details:

• API Specs: https://metmuseum.github.io/
• GH repo with CSV: https://github.com/metmuseum/openaccess

These data will need a bit of cleaning and munging to prepare for an eventual application, but there are quite a few concents that we can illustrate effectively here. Such as executing APIs on demand, hosting data in a database, custom theming, etc.

[rpodcast]

I've spent a fair bit of time reviewing the MET art collection API and associated CSV of the data. I'm still working through the best ways to clean the data to make it easier for my workshop apps, but I'm thinking the app(s) at a high level could answer a few of the following questions and let the users drill down certain aspects of the data:

Artwork Questions / Summaries

• Summary of the most frequent mediums used (materials used to create artwork). This is free text with many distinct values, so I will need to do some categorization to make it feasible.
• Distribution of artwork by countries and possible culture
• Time course of the years art were created categorized by their classification (Prints, Photographs, etc). Need to determine how to handle entries with multiple values separated by |

Artist Questions / Summaries

• Which artists have the most artwork?
• Summary of artist's country of origin
• Age of artists at time of their artwork being created/released

General App Requirements

• Implement drill-down cross-talk between plots and table summaries
• Idea: Show random selection of art and let the user rate the art for how much they like it (or a simple like/dislike). The backend of the app can record the responses and we can have fun by showing a leaderboard of the most liked or highest rated pieces.
• Show a gallery of art by type / classification and let the user choose which ones to put in their virtual museum. Somewhat related to the above idea.

[rpodcast]

Update on data exploration: I have filtered for only public domain records and those art with only a single artist (i.e. entries that do not have multiple artists split by |). We still have almost 200k rows to play with, which is good enough for me!