Question

I need to create a pivot table in R that will have text values instead of aggregated SUM or AVERAGE functions. Something like in the example below:

Original table (in a dataframe with fields User, Field, Text):

User | Field | Text

Group1 | Field 1 | Text content A
Group1 | Field 2 | Text content B
Group2 | Field 1 | Text content C
Group2 | Field 2 | Text content D
Group3 | Field 1 | Text content E

The result that I need in R using cross tabulation:

User | Field 1 | Field 2

Group1 | Text content A | Text content B
Group2 | Text content C |Text content D
Group3 | Text content E |NA
I would then need to save the result as CSV… but that’s the easy part:)

Is there any way of achieving this?

Many thanks,
MA

Answer 1

This is what reshape (going from ‘long’ format to ‘wide’ format). Other people prefer the reshape package, but both will do for this simple case.

Here’s a solution with the base function:

tmp<-data.frame(user=c("G1", "G1", "G2", "G2", "G3"), field=c("F1", "F2", "F1", "F2", "F1"), txt=c("A", "B", "C", "D", "E"))
reshape(data=tmp, direction="wide", timevar="field", idvar="user")