Documentation for check_model()

[strengejacke]

@bwiernik @IndrajeetPatil @DominiqueMakowski @mattansb I think a very common question is how to interprete these plots. We already have some more details in the docs, but probably we should provide some guidelines about the meaning of these plots in the docs:

https://easystats.github.io/performance/reference/check_model.html

Originally posted by @strengejacke in easystats/see#211 (comment)

[bwiernik]

We should split the docs up by model type like with parameters

[strengejacke]

Maybe:

1. Generic plots (seen for all model types)
2. specific plots linear/...
3. specific plots glm/...
4. specific plots mixed models/...

?

There is some overlap between those types.

[mattansb]

I think perhaps a vignette would be better than r-docs, what do you think? With a section for linear, count, binomial, mixed, etc...

[bwiernik]

I would prefer them to be organized around model type, rather than type of plot, because the same plot might be used for different things for different models (eg, a scale-location plot is used to check homogeneity in a linear model, but overdispersion in a count model)

[mattansb]

Agreed.

[strengejacke]

ok, makes sense.

[strengejacke]

Anyone who is willing to make a first draft?

[strengejacke]

(bump)

[strengejacke]

(bump)

[bwiernik]

Will start on it this weekend

[strengejacke]

bump

[strengejacke]

I'd like to submit performance soon, do you think we can add some stuff to the docs?

[bwiernik]

I'm writing a draft of this vignette on my flight today.

[IndrajeetPatil]

bump

[strengejacke]

I hope you landed safely, Brenton. :-D
Can you look at https://easystats.github.io/performance/reference/check_model.html and maybe add/revise there? It's probably faster for now than writing a whole vignette.