show factor levels in `tidy_add_variable_labels`?

[aghaynes]

When you have factors in the model, you need to know which level of the factor the specific term refers too... Would it be possible to extract the factor level (perhaps from term) and put it in a level_label variable or something like that?

Something like...?

m <- lm(mpg ~ f_cyl, mtcars)
tidy_and_attach(m) |> 
  tidy_add_variable_labels() |> 
  mutate(level_label = if_else(var_type %in% c("categorical", "dichotomous"), 
                               stringr::str_replace(term, variable, ""), "")) |> 
  select(term:var_label, var_type, level_label)
# A tibble: 3 × 5
  term        variable    var_label   var_type    level_label
  <chr>       <chr>       <chr>       <chr>       <chr>      
1 (Intercept) (Intercept) (Intercept) intercept   ""         
2 f_cyl6      f_cyl       N cylinders categorical "6"        
3 f_cyl8      f_cyl       N cylinders categorical "8"        

This would certainly be useful in tidy_add_variable_labels. I'm not sure about other functions...

[larmarange]

The purpose of tidy_add_variable_labels() is and only is to add the appropriate variable labels. If it was not called before in the pipeline, tidy_add_variable_labels() calls tidy_identify_variables() whose purpose is to identify the corresponding variable of each term.

Regarding extracting the factor level, it is the purpose of tidy_add_term_labels().

You also have tidy_add_reference_rows() to add a row with the reference level.

In broom.helpers, each single operation is done by a different function, and all these functions can be combined in a pipeline.

Most users do not need these individual functions and would rather use tidy_plus_plus() which combines all features of broom.helpers in one function (see documentation for the full list of arguments).

See examples below

I'm closing the issue as the feature you refer to is already implemented. But feel free to reopen if you identify something missing.

library(broom.helpers)
mtcars$f_cyl <- factor(mtcars$cyl)
m <- lm(mpg ~ f_cyl, mtcars)

tidy_and_attach(m) |> 
  tidy_add_variable_labels() 
#> # A tibble: 3 × 12
#>   term      variable var_label var_class var_type var_nlevels estimate std.error
#>   <chr>     <chr>    <chr>     <chr>     <chr>          <int>    <dbl>     <dbl>
#> 1 (Interce… (Interc… (Interce… <NA>      interce…          NA    26.7      0.972
#> 2 f_cyl6    f_cyl    f_cyl     factor    categor…           3    -6.92     1.56 
#> 3 f_cyl8    f_cyl    f_cyl     factor    categor…           3   -11.6      1.30 
#> # ℹ 4 more variables: statistic <dbl>, p.value <dbl>, conf.low <dbl>,
#> #   conf.high <dbl>

tidy_and_attach(m) |> 
  tidy_add_term_labels()
#> # A tibble: 3 × 15
#>   term        variable    var_label   var_class var_type   var_nlevels contrasts
#>   <chr>       <chr>       <chr>       <chr>     <chr>            <int> <chr>    
#> 1 (Intercept) (Intercept) (Intercept) <NA>      intercept           NA <NA>     
#> 2 f_cyl6      f_cyl       f_cyl       factor    categoric…           3 contr.tr…
#> 3 f_cyl8      f_cyl       f_cyl       factor    categoric…           3 contr.tr…
#> # ℹ 8 more variables: contrasts_type <chr>, label <chr>, estimate <dbl>,
#> #   std.error <dbl>, statistic <dbl>, p.value <dbl>, conf.low <dbl>,
#> #   conf.high <dbl>

tidy_and_attach(m) |> 
  tidy_add_reference_rows() |> 
  tidy_add_term_labels()
#> # A tibble: 4 × 16
#>   term        variable    var_label   var_class var_type   var_nlevels contrasts
#>   <chr>       <chr>       <chr>       <chr>     <chr>            <int> <chr>    
#> 1 (Intercept) (Intercept) (Intercept) <NA>      intercept           NA <NA>     
#> 2 f_cyl4      f_cyl       f_cyl       factor    categoric…           3 contr.tr…
#> 3 f_cyl6      f_cyl       f_cyl       factor    categoric…           3 contr.tr…
#> 4 f_cyl8      f_cyl       f_cyl       factor    categoric…           3 contr.tr…
#> # ℹ 9 more variables: contrasts_type <chr>, reference_row <lgl>, label <chr>,
#> #   estimate <dbl>, std.error <dbl>, statistic <dbl>, p.value <dbl>,
#> #   conf.low <dbl>, conf.high <dbl>

tidy_plus_plus(m)
#> # A tibble: 3 × 17
#>   term   variable var_label var_class var_type    var_nlevels contrasts      
#>   <chr>  <chr>    <chr>     <chr>     <chr>             <int> <chr>          
#> 1 f_cyl4 f_cyl    f_cyl     factor    categorical           3 contr.treatment
#> 2 f_cyl6 f_cyl    f_cyl     factor    categorical           3 contr.treatment
#> 3 f_cyl8 f_cyl    f_cyl     factor    categorical           3 contr.treatment
#> # ℹ 10 more variables: contrasts_type <chr>, reference_row <lgl>, label <chr>,
#> #   n_obs <dbl>, estimate <dbl>, std.error <dbl>, statistic <dbl>,
#> #   p.value <dbl>, conf.low <dbl>, conf.high <dbl>

^{Created on 2023-03-31 with reprex v2.0.2}

[larmarange]

Same examples but with full tables visible. The levels are visible in the label column.

library(broom.helpers)
mtcars$f_cyl <- factor(mtcars$cyl)
m <- lm(mpg ~ f_cyl, mtcars)

tidy_and_attach(m) |> 
  tidy_add_variable_labels() |> 
  knitr::kable()

term	variable	var_label	var_class	var_type	var_nlevels	estimate	std.error	statistic	p.value	conf.low	conf.high
(Intercept)	(Intercept)	(Intercept)	NA	intercept	NA	26.663636	0.9718008	27.437347	0.0000000	24.67608	28.651192
f_cyl6	f_cyl	f_cyl	factor	categorical	3	-6.920779	1.5583482	-4.441099	0.0001195	-10.10796	-3.733599
f_cyl8	f_cyl	f_cyl	factor	categorical	3	-11.563636	1.2986235	-8.904533	0.0000000	-14.21962	-8.907653

tidy_and_attach(m) |> 
  tidy_add_term_labels() |> 
  knitr::kable()

term	variable	var_label	var_class	var_type	var_nlevels	contrasts	contrasts_type	label	estimate	std.error	statistic	p.value	conf.low	conf.high
(Intercept)	(Intercept)	(Intercept)	NA	intercept	NA	NA	NA	(Intercept)	26.663636	0.9718008	27.437347	0.0000000	24.67608	28.651192
f_cyl6	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	6	-6.920779	1.5583482	-4.441099	0.0001195	-10.10796	-3.733599
f_cyl8	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	8	-11.563636	1.2986235	-8.904533	0.0000000	-14.21962	-8.907653

tidy_and_attach(m) |> 
  tidy_add_reference_rows() |> 
  tidy_add_term_labels() |> 
  knitr::kable()

term	variable	var_label	var_class	var_type	var_nlevels	contrasts	contrasts_type	reference_row	label	estimate	std.error	statistic	p.value	conf.low	conf.high
(Intercept)	(Intercept)	(Intercept)	NA	intercept	NA	NA	NA	NA	(Intercept)	26.663636	0.9718008	27.437347	0.0000000	24.67608	28.651192
f_cyl4	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	TRUE	4	NA	NA	NA	NA	NA	NA
f_cyl6	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	FALSE	6	-6.920779	1.5583482	-4.441099	0.0001195	-10.10796	-3.733599
f_cyl8	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	FALSE	8	-11.563636	1.2986235	-8.904533	0.0000000	-14.21962	-8.907653

tidy_plus_plus(m) |> 
  knitr::kable()

term	variable	var_label	var_class	var_type	var_nlevels	contrasts	contrasts_type	reference_row	label	n_obs	estimate	std.error	statistic	p.value	conf.low	conf.high
f_cyl4	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	TRUE	4	11	0.000000	NA	NA	NA	NA	NA
f_cyl6	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	FALSE	6	7	-6.920779	1.558348	-4.441099	0.0001195	-10.10796	-3.733599
f_cyl8	f_cyl	f_cyl	factor	categorical	3	contr.treatment	treatment	FALSE	8	14	-11.563636	1.298623	-8.904533	0.0000000	-14.21962	-8.907653

^{Created on 2023-03-31 with reprex v2.0.2}

[aghaynes]

perfect, thanks!