[Feature] report() for BayesFactor objects

[DominiqueMakowski]

library(report)

rez <- BayesFactor::ttestBF(iris$Sepal.Width, iris$Sepal.Length)
report_statistics(rez)
#> [1] "BF10 > 1000"
report(rez, h0="the null hypothesis", h1="the alternative")
#> [1] "There is extreme evidence in favour of the alternative over the null hypothesis (BF10 > 1000)."

^{Created on 2024-03-28 with reprex v2.0.2}

[mattansb]

I really dislike that the default behavior of format_bf() is to truncate BFs like this... format_bf(exact = TRUE) is much better :/

library(report)

rez <- BayesFactor::ttestBF(iris$Sepal.Width, iris$Sepal.Length)
report_statistics(rez, exact = TRUE)
#> [1] "BF10 = 6.11e+107"
report(rez, exact = TRUE)
#> [1] "There is extreme evidence in favour of H1 over H0 (BF10 = 6.11e+107)."

Also works now (but can probably be better…)

rez <- BayesFactor::correlationBF(iris$Sepal.Width, iris$Sepal.Length, nullInterval = c(-0.1, 0.1))
report_statistics(rez, exact = TRUE)
#> Multiple `BFBayesFactor` models detected - reporting for the first
#>   numerator model.
#>   See help("get_parameters", package = "insight").
#> [1] "BF10 = 1.18"
report(rez, exact = TRUE)
#> Multiple `BFBayesFactor` models detected - reporting for the first
#>   numerator model.
#>   See help("get_parameters", package = "insight").
#> [1] "There is anecdotal evidence in favour of H1 over H0 (BF10 = 1.18)."

Also auto passes to bayestestR::bf_models() |> report() for linear models

rez <- BayesFactor::lmBF(mpg ~ hp, data = mtcars)
report(rez)
#> Bayes factors were computed with the `BayesFactor` package, using JZS priors.
#> Compared to the (Intercept only) model (the least supported model), we found
#> extreme evidence (BF = 5.70e+04) in favour of the hp model (the most supported
#> model).

^{Created on 2024-03-28 with reprex v2.1.0}

[DominiqueMakowski]

nice addition of the LM case!

default format

We can exact=TRUE in the example so that people are nudged to use it?

[mattansb]

Yes please!

[rempsyc]

Thanks, but we see:

── R CMD check results ───────────────────────────────────── report 0.5.8.1 ────
Duration: 3m 47.5s

❯ checking Rd \usage sections ... WARNING
  Undocumented arguments in documentation object 'report.BFBayesFactor'
    ‘...’ ‘table’
  
  Functions with \usage entries need to have the appropriate \alias
  entries, and all their arguments documented.
  The \usage entries must correspond to syntactically valid R code.
  See chapter ‘Writing R documentation files’ in the ‘Writing R
  Extensions’ manual.

❯ checking for unstated dependencies in examples ... WARNING
  '::' or ':::' import not declared from: ‘BayesFactor’
  'library' or 'require' call not declared from: ‘BayesFactor’

Error in `build_reference_index()`:
! All topics must be included in reference index
✖ Missing topics: report.BFBayesFactor
ℹ Either add to _pkgdown.yml or use @keywords internal

[rempsyc]

• Error in build_reference_index() ✅
• Some files don't follow the style guide ✅
• Lints ✅

Snapshots are still failing, but I'm afraid we'll have to fix them in a different PR :/

[rempsyc]

Snapshots updated, closes #422

[rempsyc]

Cannot fix this .01 snapshot value difference between local and GHA, so skipping it...

Error: Error: Snapshot of code has changed (variant 'windows'):
old vs new
  Code
    report_performance(x7)
  Output
-   The model's explanatory power is substantial (R2 = 0.83, 95% CI [0.79, 0.86],
+   The model's explanatory power is substantial (R2 = 0.83, 95% CI [0.79, 0.87],
    adj. R2 = 0.83) and the part related to the fixed effects alone (marginal R2)
    is of 0.95 (95% CI [0.93, 0.97])