🪦 dv design 2025 spring

Amelia

this is just a test. creating a set removes all duplicates, thus i can see how unique the ID is: it covers 100k songs out of 114k (almost good)

Angie Hjort

if you were using song names in order to avoid duplicates, you ought to do it separately instead

update: i did that above in the first cell now

added a unique ID to use in graph node.id, link.source and link.target

you were using song names as IDs, but this is not helpful because song names often repeat in the dataset (search for "bulletproof" for example)

thus your selection would then highlight wrong songs when matching just by song name

thus, i added also artist and genre fields into the ID. it's still not unique though, but it's almost unique. the dataset originally had true unique IDs

you were checking songs with selectionSet.has(node.id), but selection is an array and it doesn't have a "has" method. here i created a set based on the IDs of songs and used that below