tests/testthat/test_extract_model_info.R

In manymodelr: Build and Tune Several Models

test_that(desc= "extract_model_info works as expected",
                    code= {
                      skip_on_oldrel()
            lm_model <- fit_model(yields, "height","weight","lm")
            aov_model <- fit_model(yields, "height","weight","aov")
            # Still need to figure out the best way to expect an equal value
            # say p values, r2, etc.
            # for now:
            expect_error(extract_model_info(lm_model))
            expect_error(extract_model_info(aov_model))
            expect_error(extract_model_info(aov_model,
                                            "nope"))
            expect_error(extract_model_info(fm2))
            expect_error(extract_model_info(fm2,"gibberish"))



          expect_equal(nrow(extract_model_info(aov_model,"msq")), 2)

          expect_equal(extract_model_info(aov_model,"predictors"),
                       "weight")
          expect_type(extract_model_info(lm_model,"resids"),"double")
          expect_type(extract_model_info(lm_model,"residuals"),"double")
          expect_equal(length(extract_model_info(aov_model,c("ssq","msq"))),2)
          # lme models
           # this is purely demonstrative
          data("yields", package = "manymodelr")
          fm2 <- lmer(weight ~ height + (1 | normal), data=yields)
          expect_equal(extract_model_info(fm2,"random_groups")[[1]],2)


          expect_equal(length(extract_model_info(fm2, c("reml","log_lik"))),2)


                    })