[R] Are melt and dcast with ffdf possible?

[Ken Nussear]

Having trouble getting melt and dcast to work with ffdf

 I'm trying to cast 3 columns to a square matrix

With a regular data.frame.....

comb <- expand.grid((1:10),(1:10))
dim(comb)
comb$Dist <- (1:100)
c.melt <- melt(comb, id.vars=c("Var1","Var2"))
dcast(c.melt,Var2~Var1)

goes from
> comb
    Var1 Var2 Dist
1      1    1    1
2      2    1    2
3      3    1    3
4      4    1    4
5      5    1    5
6      6    1    6
.....
to

> c.melt
    Var1 Var2 variable value
1      1    1     Dist     1
2      2    1     Dist     2
3      3    1     Dist     3
4      4    1     Dist     4
5      5    1     Dist     5
6      6    1     Dist     6
.....
to
> dcast(c.melt,Var1~Var2)
   Var1  1  2  3  4  5  6  7  8  9  10
1     1  1 11 21 31 41 51 61 71 81  91
2     2  2 12 22 32 42 52 62 72 82  92
3     3  3 13 23 33 43 53 63 73 83  93
4     4  4 14 24 34 44 54 64 74 84  94
5     5  5 15 25 35 45 55 65 75 85  95
6     6  6 16 26 36 46 56 66 76 86  96
7     7  7 17 27 37 47 57 67 77 87  97
8     8  8 18 28 38 48 58 68 78 88  98
9     9  9 19 29 39 49 59 69 79 89  99
10   10 10 20 30 40 50 60 70 80 90 100

no trouble.

With ffdf it breaks on the melt step
comb.ff <- expand.ffgrid(ff(1:10),ff(1:10))
dim(comb.ff)
comb.ff$Dist <- ff(1:100)


> comb.ff
ffdf (all open) dim=c(100,3), dimorder=c(1,2) row.names=NULL
ffdf virtual mapping
     PhysicalName VirtualVmode PhysicalVmode  AsIs VirtualIsMatrix
PhysicalIsMatrix PhysicalElementNo PhysicalFirstCol PhysicalLastCol
PhysicalIsOpen
Var1         Var1      integer       integer FALSE          
FALSE            FALSE                 1                1              
1           TRUE
Var2         Var2      integer       integer FALSE          
FALSE            FALSE                 2                1              
1           TRUE
Dist         Dist      integer       integer FALSE          
FALSE            FALSE                 3                1              
1           TRUE
ffdf data
    Var1 Var2 Dist
1      1    1    1
2      2    1    2
3      3    1    3
4      4    1    4
5      5    1    5
6      6    1    6
....
c.melt.ff <- melt(comb.ff, id.vars=c("Var1","Var2"))
> c.melt.ff
    value NA  NA
1       1  1   1
2       2  1   2
3       3  1   3
4       4  1   4
5       5  1   5
6       6  1   6
7       7  1   7
.....

Is it possible to use this method with ffdf?

I appreciate any help

Ken