Home / Questions / 354255
Accepted
Juan Luis
Asked: 2020-05-11 03:43:41 +0800 CST 2020-05-11 03:43:41 +0800 CST

Customize table generated with sjPlot in Rmarkdown

  • 772

I am estimating two binomial logit models and I want to present the results in a stylized table . The code of the models are the following:

library(readxl)
datos_modelo <- read_excel("datos_def_excel")

library(glm2)
modelo_logit_viv <- glm2( SAP ~sexo + edad + IMC + niv_est, 
                         data = datos_modelo, family = binomial(link = "logit"),na.action = "na.omit")

modelo_logit <- glm2(SAP ~ sexo + edad + IMC + niv_est,
                    data = modelo_logit_viv$model, family = binomial(link = "logit"),na.action = "na.omit")

(I transform the variables that interest me to type factor and create the list etiqs_modelto change the name of the labels that appear in the table that I generate below)

library(sjPlot)
tab_model(modelo_logit,modelo_logit_viv,transform = "exp", title = "Salud autopercibida",pred.labels = etiqs_model, wrap.labels = 10,
           string.pred = "Variables",dv.labels = c("Modelo I","Modelo II"),show.se = TRUE,show.r2 = TRUE,
           string.est = "ODD-R", p.style = "numeric", string.se = "Er. estánd.",show.ci = FALSE)

In this example both models have the same variables, but in my work the variables differ and that is why I generate a table with two models . This generates the following table: Generated table

My questions are:

  • How can I increase the spacing between columns so that the text in the first column doesn't appear so close together? If you look closely, the text "ODD-R" appears too close to "Standard Er."

  • How can I add a vertical line or a larger separation between the results of Model I and Model II?

From what I've searched, I think I have to add the argument list=CSS(...)in tab_modelbut I don't know the parameters that I should include in list=CSS(...). The biggest help I've found on this is this .

r

1 Answers

  1. Best Answer

    Indeed, the way to adjust the appearance of the tables is by using css. A basic example:

    library(sjPlot)
data(efc)
m1 <- lm(barthtot ~ c160age + c12hour + c161sex + c172code, data = efc)
m2 <- lm(barthtot ~ nur_pst + c172code + e42dep, data = efc)
tab <- tab_model(m1, m2)
tab

    Result:

    enter image description here

    What you are looking for:

    • Add something paddingto a particular column. In this example we will do it on the first Estimatesof the first model, it is the second column
    • Add a vertical border between the two models, adding borderto the fourth column.

    The first thing is to use cat(tab$page.style)to understand what styles we can configure for our table, in this example, we will use: .col2and .col4, and the idea is to do something like this:

    CSS <- list(
  css.col2 = 'padding-right: 50px;',
  css.col4 = 'border-right: 1px solid black'
)

tab <- tab_model(m1, m2, CSS = CSS) 
tab

    Result:

    enter image description here

    Additional documentation that may be useful to you:

    • 1