Tidy characteristics data
tibble_one.Rd
Table one is a tabular description of characteristics, e.g., demographics of patients in a clinical trial, presented overall and also stratified by a categorical variable, e.g. treatment group.
tibble_one(data, formula = NULL, meta_data = NULL, row_vars = NULL, strat = NULL, by = NULL, specs_table_vals = NULL, specs_table_tests = NULL, include_pval = FALSE, expand_binary_catgs = FALSE, include_freq = FALSE, add_perc_to_cats = TRUE)
Arguments
| data | a data frame |
|---|---|
| formula | an optional formula object. The left hand side of the formula should be blank. The right hand side of the formula should contain row variables for the table. The '|' symbol can be used to include stratifying variables. If this option is used, no more than two stratifying variables should be used, and they must be separated by a * symbol. If formula is used, the strat, by, and row_vars inputs are ignored. |
| meta_data | a meta data frame. If unspecified, a meta data frame
will be created using |
| row_vars | a character vector indicating column names of row variables in the table. If unspecified, all columns are used. |
| strat | a character value indicating the column name in data that will be used to stratify the table |
| by | a character value indicating the column name in data that will be used to split the table into groups, prior to stratification. |
| specs_table_vals | named vector of character values. Names should be variables, while values should be specs. Valid specs are 'mean' and 'median' (see examples). |
| specs_table_tests | named vector of character values. Names should be variables, while values should be specs. Valid specs are 'params' or 'noparm' (see examples). |
| include_pval | T/F, should the table include a column for p-values? If p-values are included, factor variables are handled using chi-square tests, continuous variables are handled using t-tests or ANOVA, depending on the number of categories in the table stratification. |
| expand_binary_catgs | T/F, should all categories be included for binary categorical variables? (This only applies to binary variables.) |
| include_freq | T/F, should frequency values be included for categorical variables? |
| add_perc_to_cats | T/F, should categorical variable labels be appended with a percent sign? |
Value
a tibble containing summary values
that describe characteristics of observations in data , which can
subsequently be sent to different modes of output
(see to_word and to_kable).
Examples
data("pbc_tbl1") # report median albumin instead of mean # use kruskal wallis test for albumin tibble_one( pbc_tbl1, formula = ~ . | trt, include_freq = FALSE, include_pval = TRUE, specs_table_vals = c(albumin = 'median'), specs_table_tests = c(albumin = 'nopars') )#> # A tibble: 19 x 7 #> group variable labels Overall `D-penicillmain` Placebo `P-value` #> <fct> <fct> <chr> <chr> <chr> <chr> <chr> #> 1 None descr No. of observ~ 418 158 154 "" #> 2 None age Age 50.7 (10.~ 51.4 (11.0) 48.6 (9.~ 0.018 #> 3 None sex Female, % 89.5 86.7 90.3 0.421 #> 4 None status Status, % "" "" "" 0.894 #> 5 None status Censored 55.5 52.5 55.2 "" #> 6 None status Transplant 5.98 6.33 5.84 "" #> 7 None status Dead 38.5 41.1 39.0 "" #> 8 None stage Stage, % "" "" "" 0.201 #> 9 None stage One 5.10 7.59 2.60 "" #> 10 None stage Two 22.3 22.2 20.8 "" #> 11 None stage Three 37.6 35.4 41.6 "" #> 12 None stage Four 35.0 34.8 35.1 "" #> 13 None ascites Ascites, % 7.69 8.86 6.49 0.567 #> 14 None bili Bili 3.22 (4.4~ 2.87 (3.63) 3.65 (5.~ 0.133 #> 15 None edema Edema, % "" "" "" 0.877 #> 16 None edema None 84.7 83.5 85.1 "" #> 17 None edema Untreated or ~ 10.5 10.1 8.44 "" #> 18 None edema Treatment res~ 4.78 6.33 6.49 "" #> 19 None albumin Albumin 3.53 [3.2~ 3.56 [3.21-3.83] 3.54 [3.~ 0.874