R : Create and transform a dataframe based on existing dataframe - r

Question

My original question was poorly defined and confusing. Here it is again with extraneous columns removed for clarity and additional background.
My end game is to create a forced network graph using networkD3. To do this I need two dataframes.
The first dataframe (dfNodes) lists each node in the graph and its grouping. Volt, Miata and Prius are cars, so they get the group value of 1.
Crusader is a bus so it is in group 3. Etc.
dfNodes dataframe:
id name group
0 vehicle 0
1 car 1
2 truck 2
3 bus 3
4 volt 1
5 miata 1
6 prius 1
7 tahoe 2
8 suburban 2
9 crusader 3
From this dataframe I need to construct the dataframe dfLinks to provide the linkages between nodes. It must look like:
dfLinks dataframe:
source target
0 4
0 5
0 6
0 7
0 8
0 9
1 4
1 5
1 6
2 7
2 8
3 9
This shows the following:
vehicle is linked to volt, miata, prius, tahoe, suburban, crusader (they are all vehicles). (0, 4; 0, 5...0, 9)
car is linked to volt, miata, prius (1, 4; 1, 5; 1, 6)
truck is linked to tahoe, suburban (2, 7 ; 2, 8)
bus is linked to crusader (3, 9)
It may appear strange to link vehicle-->model name (volt...crusader)--> type (car/bus/truck) instead of
vehicle--> type --> model name
but that is the form that I need for my graph.

Answer 1

I don't understand the criteria for creating the second dataframe; for example 'car' appears three times in the second data frame but is present four times in the first data frame? If you can provide more precise details about the transformation I can try and help more, but my first suggestion would be to look at the melt() and cast() functions in the reshape package:
http://www.statmethods.net/management/reshape.html

Answer 2

Based on the replies from Phil and aosmith I revisited my question again. I have the luxury of creating the source dataframe (dfNodes) so I approached the problem from further upstream, first by creating dfNodes by combining dataframes, then creating dfLinks independently from dfNodes instead of based on dfNodes. I feel my approach is very much a kludge but it does give me the result I am looking for. Here is the code. Other approaches, advice and criticism welcomed! As you can tell, I am new to R.
# Separate dataframes for the different categories of data
vehicle <- data.frame(source=0, name="vehicle", group=0)
carGroup <- data.frame(source=1, name="car", group=1)
truckGroup <- data.frame(source=2, name="truck", group=2)
busGroup <- data.frame(source=3, name="bus", group=3)
cars <- data.frame(source=c(4,5,6),
name=c("volt", "miata", "prius"),
group=c(1,1,1))
trucks <- data.frame(source=c(7,8),
name=c("tahoe", "suburban"),
group=c(2,2))
buses <- data.frame(source=9,
name="crusader",
group=3)
# 1. Build the dfNodes dataframe
dfNodes <- rbind(vehicle, carGroup, truckGroup, busGroup, cars, trucks, buses)
names(dfNodes)[1]<-"id"
dfNodes
# 2. Build dfLinks dataframe. Only source and target columns needed
# Bind Vehicle to the 3 different vehicle dataframes
vehicleToCars <- cbind(vehicle, cars)
vehicleToTrucks <- cbind(vehicle, trucks)
vehicleToBuses <- cbind(vehicle, buses)
# Bind the Vehicle Groups to the vehicles
carGroupToCars <- cbind(carGroup, cars)
truckGroupToTrucks <- cbind(truckGroup, trucks)
busGroupToBuses <- cbind(busGroup, buses)
# Stack into the final dfLinks dataframe
dfLinks <- rbind(vehicleToCars, vehicleToTrucks, vehicleToBuses,
carGroupToCars, truckGroupToTrucks, busGroupToBuses)
names(dfLinks)[4]<-"target"
dfLinks