search

probmods / probmods2

probmods 2: electric boogaloo
https://probmods.org/v2
296 stars 76 forks source link

Clarify model criticism section in bayesian data analysis chapter #127

Closed marekjg closed 3 years ago

marekjg commented 3 years ago

Hi, I'm going through the book and the first thing I've encountered that seems a bit confusing is the part where we split model prediction for each outcome, tournament and pattern:

var itemInfo = {pattern: pattern, tournament: tournament, outcome: outcome}
var itemData = _.filter(towData, itemInfo)
var predicted_y = b0 + itemData[0]["nWins"]*b1

While itemData contains ~30 entries, with different participants ID, we take itemData[0]["nWins"] as a representative. I think that clarifying what pattern means would be beneficial for comprehension, because from the inference side everything is clear.

ngoodman commented 3 years ago

thanks! that chapter has been pretty completely revised, so i'm hoping this is not an issue any more.