Export a list of matrices nicely to the same worksheet in Excel

Question

I have a list of 4 matrices. The first two of them:

$`1857-1903`
                      Bank.o.kassa Obligationer       Lån    Aktier Placeringar.andra.ftg Fodringar Reala.tillgångar Övriga.tillgångar
Bank.o.kassa             1.0000000    0.8014382 0.8079718 0.4421687                    NA 0.8395201        0.9144023         0.7995480
Obligationer             0.8014382    1.0000000 0.8422192 0.4590398                    NA 0.7871074        0.8815054         0.7003110
Lån                      0.8079718    0.8422192 1.0000000 0.5476716                    NA 0.8932723        0.9295665         0.7846206
Aktier                   0.4421687    0.4590398 0.5476716 1.0000000                    NA 0.7203681        0.4221815         0.6281095
Placeringar.andra.ftg           NA           NA        NA        NA                     1        NA               NA                NA
Fodringar                0.8395201    0.7871074 0.8932723 0.7203681                    NA 1.0000000        0.8869905         0.9395209
Reala.tillgångar         0.9144023    0.8815054 0.9295665 0.4221815                    NA 0.8869905        1.0000000         0.8158413
Övriga.tillgångar        0.7995480    0.7003110 0.7846206 0.6281095                    NA 0.9395209        0.8158413         1.0000000
Kapitalinkomster         0.9163688    0.8960790 0.9318508 0.4985449                    NA 0.9277740        0.9821501         0.8755644
                      Kapitalinkomster
Bank.o.kassa                 0.9163688
Obligationer                 0.8960790
Lån                          0.9318508
Aktier                       0.4985449
Placeringar.andra.ftg               NA
Fodringar                    0.9277740
Reala.tillgångar             0.9821501
Övriga.tillgångar            0.8755644
Kapitalinkomster             1.0000000

$`1904-1948`
                      Bank.o.kassa Obligationer         Lån       Aktier Placeringar.andra.ftg   Fodringar Reala.tillgångar Övriga.tillgångar
Bank.o.kassa             1.0000000   -0.5636371  0.12902979 -0.658883574          -0.202688601  0.76474140       0.81544129         0.4642105
Obligationer            -0.5636371    1.0000000  0.23820065  0.369712399           0.649156560 -0.69097442      -0.67090178        -0.3108845
Lån                      0.1290298    0.2382006  1.00000000 -0.301682319           0.663984986 -0.07873989       0.00901089         0.2894697
Aktier                  -0.6588836    0.3697124 -0.30168232  1.000000000          -0.005049947 -0.69634496      -0.77966741        -0.7496402
Placeringar.andra.ftg   -0.2026886    0.6491566  0.66398499 -0.005049947           1.000000000 -0.44833419      -0.31431360         0.1089602
Fodringar                0.7647414   -0.6909744 -0.07873989 -0.696344963          -0.448334185  1.00000000       0.84590718         0.5765548
Reala.tillgångar         0.8154413   -0.6709018  0.00901089 -0.779667406          -0.314313596  0.84590718       1.00000000         0.5367746
Övriga.tillgångar        0.4642105   -0.3108845  0.28946968 -0.749640198           0.108960190  0.57655477       0.53677459         1.0000000
Kapitalinkomster        -0.1114329    0.3693150  0.38160001 -0.125296598           0.720230427 -0.17280772      -0.13946215         0.3720201
                      Kapitalinkomster
Bank.o.kassa                -0.1114329
Obligationer                 0.3693150
Lån                          0.3816000
Aktier                      -0.1252966
Placeringar.andra.ftg        0.7202304
Fodringar                   -0.1728077
Reala.tillgångar            -0.1394622
Övriga.tillgångar            0.3720201
Kapitalinkomster             1.0000000

I would like to export them to the same worksheet in Excel. The problem is that if I, example, do write.csv2(my.list,file="my.list.csv2") the matrices are not separated, there is no spacing between them. On the other hand I know how to use XLConnect to export my list to a workbook with several sheets (4 in this case). But I want my matrices to be in the same worksheet, separated by some spacing, with listnames (i.e. $1857-1903) present. And it also would be nice if rownames would be present... Is it possible? Could not find any answer. Best Regards!

Edit:

I accepted @January answer. Thanks! But if I use write.csv2 instead of write.table (in order to get the desired output) I get a warning message:

In write.csv2(export, file = "funkcorr.csv", quote = F, sep = ",") : attempt to set 'sep' ignored. Then, the numbers in Excel are with . instead of ,. So, I need to manually convert . to , in Excel. Any idea why sep is ignored?

Answer 1

The following is an XLConnect-based solution:

# Preparation of dummy data
data = rep(list(WorldPhones), 5)
names(data) = LETTERS[1:5]

require(XLConnect)
wb = loadWorkbook("matrix.xlsx", create = TRUE)
# Create a new sheet
createSheet(wb, name = "mysheet")
# cumulative length (rows) of matrices
# +2 = 1 for list names, 1 for header row
cumlen = cumsum(c(1, head(sapply(data, nrow), n = -1) + 2))
# Write data rows (implicitly vectorized!)
writeWorksheet(wb, data = data, sheet = "mysheet", startRow = cumlen + 1, header = TRUE)
# Write list names
writeWorksheet(wb, data = as.list(names(data)), sheet = "mysheet", startRow = cumlen, header = FALSE)
saveWorkbook(wb)

Answer 2

Assuming all matrices have the same number of columns, and that the list is called matrices.l:

export <- NULL
nc <- ncol( matrices.l[[1]] )

for( n in names( matrices.l ) ) 
  export <- rbind( export, c( n, rep( "", nc - 1 ) ), matrices.l[[n]], "" )

write.table( export, file= ...,  quote= F, sep= ",", ... )

This is not elegant and makes the cardinal sin of creating an R object in a for loop, but for four matrices will do.

Answer 3

Looking into the problem a little more, I still think you can get something workable with sink(), but I'm not sure if there are better options. Here's a basic example:

Example data

set.seed(1)
myList <- list(matrixA = matrix(sample(300, 60), nrow = 10,
                                dimnames = list(LETTERS[1:10], 
                                                paste0("V", 1:6))),
               matrixB = matrix(sample(300, 50), nrow = 10,
                                dimnames = list(LETTERS[1:10], 
                                                paste0("V", 1:5))))

Using sink() and lapply()

sink("something.else.csv", type="output")
invisible(lapply(names(myList), 
                 function(x) { print(x)
                               dput(write.csv(myList[[x]])) } ))
sink()

Here's what the resulting CSV files look like:

[1] "matrixA"
"","V1","V2","V3","V4","V5","V6"
"A",80,296,262,131,214,120
"B",112,52,290,162,168,215
"C",171,198,182,133,202,109
"D",270,111,35,50,143,61
"E",60,221,74,286,136,291
"F",266,142,107,178,258,25
"G",278,204,4,210,6,78
"H",194,281,105,29,121,127
"I",184,108,237,190,185,161
"J",18,219,93,282,174,99
NULL
[1] "matrixB"
"","V1","V2","V3","V4","V5"
"A",274,297,122,65,171
"B",88,243,199,16,92
"C",137,100,112,173,70
"D",99,96,91,234,256
"E",193,298,209,208,163
"F",77,291,56,212,55
"G",141,246,195,121,33
"H",225,111,34,108,264
"I",25,220,67,213,233
"J",255,270,39,158,151
NULL

This opens just fine in any spreadsheet program, and you can easily search for all instances of "NULL" and replace it with nothing to have a blank row between each matrix. You can also, of course, use write.csv2() instead of write.csv() if that is what you need.