add_difference.tbl_svysummary.md
In gtsummary: Presentation-Ready Data Summary and Analytic Result Tables

add_difference.tbl_svysummary() snapshots of common outputs

Code
  select(as.data.frame(add_difference(tbl_svysummary(svy_trial, by = trt)),
  col_labels = FALSE), -all_stat_cols())
Output
                      label estimate    conf.low p.value
  1                     Age     0.44   -3.7, 4.6     0.8
  2                 Unknown     <NA>        <NA>    <NA>
  3    Marker Level (ng/mL)    -0.20 -0.44, 0.05    0.12
  4                 Unknown     <NA>        <NA>    <NA>
  5                 T Stage     0.12 -0.16, 0.40    <NA>
  6                      T1     <NA>        <NA>    <NA>
  7                      T2     <NA>        <NA>    <NA>
  8                      T3     <NA>        <NA>    <NA>
  9                      T4     <NA>        <NA>    <NA>
  10                  Grade     0.07 -0.20, 0.35    <NA>
  11                      I     <NA>        <NA>    <NA>
  12                     II     <NA>        <NA>    <NA>
  13                    III     <NA>        <NA>    <NA>
  14         Tumor Response    -4.2%  -17%, 8.9%     0.5
  15                Unknown     <NA>        <NA>    <NA>
  16           Patient Died    -5.8%  -20%, 8.0%     0.4
  17 Months to Death/Censor     -1.2  -2.7, 0.26    0.11

Code
  select(as.data.frame(add_difference(tbl_svysummary(svy_titanic, by = Survived)),
  col_labels = FALSE), -all_stat_cols())
Message
  The following warnings were returned during `add_difference()`:
  ! For variable `Freq` (`Survived`) and "estimate", "statistic", "p.value", "parameter", "conf.low", and "conf.high" statistics: observations with zero weight not used for calculating dispersion
  ! For variable `Freq` (`Survived`) and "estimate", "statistic", "p.value", "parameter", "conf.low", and "conf.high" statistics: observations with zero weight not used for calculating dispersion
Output
      label estimate   conf.low p.value
  1   Class     0.56 0.47, 0.65    <NA>
  2     1st     <NA>       <NA>    <NA>
  3     2nd     <NA>       <NA>    <NA>
  4     3rd     <NA>       <NA>    <NA>
  5    Crew     <NA>       <NA>    <NA>
  6     Sex     0.80 0.71, 0.89    <NA>
  7    Male     <NA>       <NA>    <NA>
  8  Female     <NA>       <NA>    <NA>
  9     Age     0.09 0.00, 0.18    <NA>
  10  Child     <NA>       <NA>    <NA>
  11  Adult     <NA>       <NA>    <NA>
  12   Freq     -323  -573, -73   0.014

add_difference.tbl_svysummary(x) messaging

Code
  add_difference(tbl_svysummary(svy_trial, include = age))
Condition
  Error in `add_difference()`:
  ! Cannot run `add_difference()` when `tbl_summary(by)` column does not have exactly two levels.

Code
  tbl <- add_difference(tbl_svysummary(svy_trial, by = trt, percent = "row",
    include = grade))
Condition
  Warning:
  The `add_difference()` results for categorical variables may not compatible with `tbl_summary(percent = c('cell', 'row'))`.
  i Use column percentages instead, `tbl_summary(percent = 'column')`.

add_difference.tbl_svysummary() + add_p.tbl_svysummary()

Code
  dplyr::select(as.data.frame(add_p(add_difference(tbl_svysummary(svy_trial, by = trt,
    include = c(age, response, grade), missing = "no"), test = ~"smd")),
  col_label = FALSE), -all_stat_cols())
Output
             label estimate    conf.low p.value
  1            Age    -0.03 -0.32, 0.25     0.7
  2 Tumor Response    -0.09 -0.37, 0.19     0.5
  3          Grade     0.07 -0.20, 0.35     0.9
  4              I     <NA>        <NA>    <NA>
  5             II     <NA>        <NA>    <NA>
  6            III     <NA>        <NA>    <NA>

add_difference.tbl_svysummary(test) messaging

Code
  add_difference(tbl_svysummary(svy_trial, by = trt, include = age), test = age ~
    letters)
Condition
  Error in `add_difference()`:
  ! Invalid value passed in `test` argument for variable "age".
  i Expecting a <string> or <function>, not a character vector.