A function that takes a single dependent variable with a vector of explanatory variable names (continuous or categorical variables) to produce a summary table.

summary_factorlist(
  .data,
  dependent = NULL,
  explanatory,
  cont = "mean",
  cont_nonpara = NULL,
  cont_cut = 5,
  cont_range = TRUE,
  p = FALSE,
  p_cont_para = "aov",
  p_cat = "chisq",
  column = TRUE,
  total_col = FALSE,
  orderbytotal = FALSE,
  digits = c(1, 1, 3, 1),
  na_include = FALSE,
  na_include_dependent = FALSE,
  na_complete_cases = FALSE,
  na_to_p = FALSE,
  fit_id = FALSE,
  add_dependent_label = FALSE,
  dependent_label_prefix = "Dependent: ",
  dependent_label_suffix = "",
  add_col_totals = FALSE,
  include_col_totals_percent = TRUE,
  col_totals_rowname = NULL,
  col_totals_prefix = "",
  add_row_totals = FALSE,
  include_row_totals_percent = TRUE,
  include_row_missing_col = TRUE,
  row_totals_colname = "Total N",
  row_missing_colname = "Missing N",
  catTest = NULL
)

Arguments

.data

Dataframe.

dependent

Character vector of length 1: name of dependent variable (2 to 5 factor levels).

explanatory

Character vector of any length: name(s) of explanatory variables.

cont

Summary for continuous explanatory variables: "mean" (standard deviation) or "median" (interquartile range). If "median" then non-parametric hypothesis test performed (see below).

cont_nonpara

Numeric vector of form e.g. c(1,2). Specify which variables to perform non-parametric hypothesis tests on and summarise with "median".

cont_cut

Numeric: number of unique values in continuous variable at which to consider it a factor.

cont_range

Logical. Median is show with 1st and 3rd quartiles.

p

Logical: Include null hypothesis statistical test.

p_cont_para

Character. Continuous variable parametric test. One of either "aov" (analysis of variance) or "t.test" for Welch two sample t-test. Note continuous non-parametric test is always Kruskal Wallis (kruskal.test) which in two-group setting is equivalent to Mann-Whitney U /Wilcoxon rank sum test.

For continous dependent and continuous explanatory, the parametric test p-value returned is for the Pearson correlation coefficient. The non-parametric equivalent is for the p-value for the Spearman correlation coefficient.

p_cat

Character. Categorical variable test. One of either "chisq" or "fisher".

column

Logical: Compute margins by column rather than row.

total_col

Logical: include a total column summing across factor levels.

orderbytotal

Logical: order final table by total column high to low.

digits

Number of digits to round to (1) mean/median, (2) standard deviation / interquartile range, (3) p-value, (4) count percentage.

na_include

Logical: make explanatory variables missing data explicit (NA).

na_include_dependent

Logical: make dependent variable missing data explicit.

na_complete_cases

Logical: include only rows with complete data.

na_to_p

Logical: include missing as group in statistical test.

fit_id

Logical: allows merging via finalfit_merge.

add_dependent_label

Add the name of the dependent label to the top left of table.

dependent_label_prefix

Add text before dependent label.

dependent_label_suffix

Add text after dependent label.

add_col_totals

Logical. Include column total n.

include_col_totals_percent

Include column percentage of total.

col_totals_rowname

Logical. Row name for column totals.

col_totals_prefix

Character. Prefix to column totals, e.g. "N=".

add_row_totals

Logical. Include row totals. Note this differs from total_col above particularly for continuous explanatory variables.

include_row_totals_percent

Include row percentage of total.

include_row_missing_col

Logical. Include missing data total for each row. Only used when add_row_totals is TRUE.

row_totals_colname

Character. Column name for row totals.

row_missing_colname

Character. Column name for missing data totals for each row.

catTest

Deprecated. See p_cat above.

Value

Returns a factorlist dataframe.

Details

This function aims to produce publication-ready summary tables for categorical or continuous dependent variables. It usually takes a categorical dependent variable to produce a cross table of counts and proportions expressed as percentages or summarised continuous explanatory variables. However, it will take a continuous dependent variable to produce mean (standard deviation) or median (interquartile range) for use with linear regression models.

See also

Examples

library(finalfit) library(dplyr) # Load example dataset, modified version of survival::colon data(colon_s) # Table 1 - Patient demographics ---- explanatory = c("age", "age.factor", "sex.factor", "obstruct.factor") dependent = "perfor.factor" colon_s %>% summary_factorlist(dependent, explanatory, p=TRUE)
#> Warning: Chi-squared approximation may be incorrect
#> label levels No Yes p #> Age (years) Mean (SD) 59.8 (11.9) 58.4 (13.3) 0.542 #> Age <40 years 68 (7.5) 2 (7.4) 1.000 #> 40-59 years 334 (37.0) 10 (37.0) #> 60+ years 500 (55.4) 15 (55.6) #> Sex Female 432 (47.9) 13 (48.1) 1.000 #> Male 470 (52.1) 14 (51.9) #> Obstruction No 715 (81.2) 17 (63.0) 0.035 #> Yes 166 (18.8) 10 (37.0)
# summary.factorlist() is also commonly used to summarise any number of # variables by an outcome variable (say dead yes/no). # Table 2 - 5 yr mortality ---- explanatory = c("age.factor", "sex.factor", "obstruct.factor", "perfor.factor") dependent = "mort_5yr" colon_s %>% summary_factorlist(dependent, explanatory)
#> Note: dependent includes missing data. These are dropped.
#> label levels Alive Died #> Age <40 years 31 (6.1) 36 (8.9) #> 40-59 years 208 (40.7) 131 (32.4) #> 60+ years 272 (53.2) 237 (58.7) #> Sex Female 243 (47.6) 194 (48.0) #> Male 268 (52.4) 210 (52.0) #> Obstruction No 408 (82.1) 312 (78.6) #> Yes 89 (17.9) 85 (21.4) #> Perforation No 497 (97.3) 391 (96.8) #> Yes 14 (2.7) 13 (3.2)