r - How to create a data.table with one row filled with NA based on another data.table -

say have following data.table (but think of table lot of columns dynamically changing names)

dt <- data.table(a=rep(1l,3),b=rep(1.32,3),d=rep("qwe",3)) dt       b   d 1: 1 1.32 qwe 2: 1 1.32 qwe 3: 1 1.32 qwe 

say want create row nas rbindlist dt.

first try (it not working)

dt1 <- dt[1][,colnames(dt):=na] dt1      b  d 1: na na na rbindlist(list(dt1,dt))          b  d 1:   na   na na 2: true true na 3: true true na 4: true true na 

this not working because dt1 casted when :=na called (something called plonking seems, because if provide full column when := rhs type overwrite lhs...)

the question then, how can extract row of data.table , fill na or create data.table, filled na have same column names , column type another

there many ways of doing it, here's one:

rbind(dt[na], dt) #       b   d #1: na   na  na #2:  1 1.32 qwe #3:  1 1.32 qwe #4:  1 1.32 qwe