I would like to transpose a set of columns in my data frame a. without needing to explicitly identifying the transposed column names in the script (there are >900) AND b. with the transposed columns being of varied data types
My data is horribly structured survey data with over 900 columns!! most columns are survey answers, and i want a very simple structure which is Name, Question, Answer
I've already tried the melt function of the reshape library however, i'm hitting the following snag noted in the the melt documentation: "..In the current implementation, there is only one assumption that melt makes: all measured values must be of the same type"
name Q1 Q2 Q3 Q900
Jill 5 Sometimes Yes 3
Eve 2 Never No 2
name Question Answer
Jill Q1 5
Jill Q2 Sometimes
Jill Q3 Yes
Jill Q900 3
Eve Q1 2
Eve Q2 Never
Eve Q3 No
Eve Q900 2
