Reputation: 1
I am using the tab_model
function from the sjPlot
package in R to do a table with 18 univariate models, all with the same response variable, but with one different variable each, as an example, this could be the first two models:
GEE1 <- geeglm(hospitalization ~ xxx,
id = record_id, data = dades, family = binomial, corstr = "ar1")
GEE2 <- geeglm(hospitalization ~ xxx,
id = record_id, data = dades, family = binomial, corstr = "ar1")
When using tab_model
, I am getting one column (or set of columns) for each model, I would like to have them all in the same set of columns . Meaning just one column for all the estimates, all the p-values etc. But now I have one set (estimates, 95% and p-value) of columns per each model. making the table, obviously too wide. I want to have just all the models listed with the results.
I have tried to do a function and use kable, but it is very manual and I would love to find an alternative
A dummie example could be:
data = mtcars
m1 = lm(mpg~ disp, data = data)
m2 = lm(mpg~ cyl, data = data)
library(sjPlot)
tab_model(m1,m2,show.intercept = FALSE)
If I run this I get a table somewhat similar to:
mpg | mpg | |||||
---|---|---|---|---|---|---|
Predictors | Estimates | $C I$ | $p$ | Estimates | $C I$ | $p$ |
disp | -0.04 | $-0.05--0.03$ | $<0.001$ | |||
cy1 | -2.88 | $-3.53--2.22$ | $<0.001$ |
I would need to collapse all in one column as I am fitting this for 18 variables (I need that to do a model selection with Hosmer-Lemeshow algoritm due to the structure of my data).
It has to be something like:
mpg | |||
---|---|---|---|
Predictors | Estimates | $C I$ | $p$ |
disp | -0.04 | $-0.05--0.03$ | $<0.001$ |
cy1 | -2.88 | $-3.53--2.22$ | $<0.001$ |
And then I wil export it to latex, tat culd be another question but if you know the answer drop it here too :)
Upvotes: 0
Views: 37