In overdodactyl/mRclwhip: Whippin' Data Into Shape at Mayo Clinic

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(mRclwhip)

The idea behind mRclwhip::polish() is to create a more 'complete product' than what is returned from broom::tidy(), for example.

The function can either return a flextable::flextable() object if the argument .flextable = TRUE.
Or, if the argument .flextable = FALSE, then a list with two objects is returned where the first object in the list is a dataframe of the polished results that can be further manipulated.

The second object in the list is a vector of row numbers that can then be passed into flextable::padding() (after passing the dataframe into mRclwhip::format_flextable() of course) to pad the rows with factor levels.

The example at the end will provide further clarification.

Example using `coxph()`

mockstudy <- arsenal::mockstudy
coxph_res <- survival::coxph(survival::Surv(fu.time, fu.stat) ~ race + sex + bmi, data = mockstudy)

Hmisc::label(mockstudy$sex) <- "Gender"

coxph_res %>%
  polish(
    .header1 = list(values = c("", "HR (95% CI)")),
    exponentiate = T
  )

Example using `lm()`

mtcars <- mtcars %>%
  dplyr::mutate_at(dplyr::vars(gear, am, carb), as.character)

mtcars <- Hmisc::upData(
  mtcars,
  labels = c(mpg = "MPG",
             hp = "HP",
             gear = "# of gears",
             am = "Automatic",
             wt = "Weight",
             carb = "Carb"
             ),
  print = F
  )

lm_res <- lm(mpg ~ hp + gear + am + wt + carb, data = mtcars)

lm_res %>%
  polish(
    .header1 = list(values = c("", "Estimate (95% CI)"))
  )

Example using `lm()` with labels argument used correctly

mtcars <- mtcars %>%
  dplyr::mutate_at(dplyr::vars(gear, am, carb), as.character)

labs <- c(hp = "HP",
          gear = "# of gears",
          am = "Automatic",
          wt = "Weight",
          carb = "Carb"
          )

lm_res <- lm(mpg ~ hp + gear + am + wt + carb, data = mtcars)

lm_res %>%
  polish(
    .header1 = list(values = c("", "Estimate (95% CI)")),
    .labels = labs
  )

Example using `lm()` with labels argument used incorrectly

mtcars <- mtcars %>%
  dplyr::mutate_at(dplyr::vars(gear, am, carb), as.character)

labs <- c(mpg = "MPG",
          hp = "HP",
          gear = "# of gears",
          am = "Automatic",
          wt = "Weight",
          carb = "Carb"
)

lm_res <- lm(mpg ~ hp + gear + am + wt + carb, data = mtcars)

lm_res %>%
  polish(
    .header1 = list(values = c("", "Estimate (95% CI)")),
    .labels = labs
  )

Example using `.flextable = FALSE` where a list with a dataframe and a vector of row numbers is returned

mtcars <- mtcars %>%
  dplyr::mutate_at(dplyr::vars(gear, am, carb), as.character)

mtcars <- Hmisc::upData(
  mtcars,
  labels = c(mpg = "MPG",
             hp = "HP",
             gear = "# of gears",
             am = "Automatic",
             wt = "Weight",
             carb = "Carb"
             ),
  print = F
  )

lm_res <- lm(mpg ~ hp + gear + am + wt + carb, data = mtcars)

res <- lm_res %>%
  polish(
    .flextable = FALSE
  )

res

Returning the dataframe allows further manipulation if desired. The row numbers returned for the rows with factor levels can then be used in flextable::padding() to add padding

res[[1]] %>% 
  mRclwhip::format_flextable(header1 = list(values = c("", "Estimate (95% CI)"))) %>% 
  flextable::padding(i = res[[2]], j = 1, padding.left = 25)