Suppose we have a data frame like the one below. The numbers are irrelevant to my actual query, so I have simply given them the same value.
df <- data.frame(ABC_1 = c(1,2,3),
ABC_2 = c(1,2,3),
ABC_3 = c(1,2,3),
ABC_4 = c(1,2,3),
ABC_5 = c(1,2,3),
ABC_6 = c(1,2,3),
ABC_7 = c(1,2,3),
ABC_8 = c(1,2,3),
ABC_9 = c(1,2,3),
ABC_10 = c(1,2,3),
ABC_11 = c(1,2,3),
ABC_12 = c(1,2,3),
ABC_13 = c(1,2,3),
ABC_14 = c(1,2,3),
ABC_15 = c(1,2,3),
ABC_16 = c(1,2,3),
ABC_17 = c(1,2,3),
ABC_18 = c(1,2,3),
ABC_19 = c(1,2,3),
ABC_20 = c(1,2,3))
Often it is the case that when one has to run a model in lavaan, they have to fit several variables into the model syntax with similar names, such as below:
mod.abc <- '
ABC_Variable =~ ABC_01 + ABC_02 + ABC_03 + ABC_04 + ABC_05 +
ABC_06 + ABC_07 + ABC_08 + ABC_09 + ABC_10 +
ABC_11 + ABC_12 + ABC_13 + ABC_14 + ABC_15 +
ABC_16 + ABC_17 + ABC_18 + ABC_19 + ABC_20
'
To generate repeating text sequences in R, one can do this easily with a function like this:
rep("ABC",20)
Which gives us a sequence of ABC's:
[1] "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC"
[12] "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC" "ABC"
However, copying and pasting these into the lavaan model syntax is still time consuming since one has to delete all the quotation marks. Is there a possibly easier way that doesn't require manually writing in the variables?
You can use:
paste("ABC_variable =~", paste(names(df), collapse = " + "))
[1] "ABC_variable =~ ABC_1 + ABC_2 + ABC_3 + ABC_4 + ABC_5 + ABC_6 + ABC_7 +
ABC_8 + ABC_9 + ABC_10 + ABC_11 + ABC_12 + ABC_13 + ABC_14 + ABC_15 + ABC_16 +
ABC_17 + ABC_18 + ABC_19 + ABC_20"
Or if I'm building more complex models, I prefer sprintf() which offers more flexibility and readability over multiple nested calls of paste().
sprintf("ABC_variable =~ %s", paste(names(df), collapse = " + "))