Hello,

It's much easier than you think, the first two columns of the input matrix are the row and column numbers into the output matrix, therefore those columns form an index matrix. Just see:

x <- scan(text="
1 1 3
1 2 54
1 3 11
1 4 17
2 1 5
2 4 78
2 5 20
")

mat <- matrix(x, ncol = 3, byrow=TRUE)

result <- matrix(0, max(mat[, 1]), max(mat[, 2]))
result[ mat[, 1:2] ] <- mat[, 3]


Easy, no?

Hope this helps,

Rui Barradas
Em 18-10-2012 13:44, Rui Esteves escreveu:
Hi,

I downloaded a dataset from UCI repositories named Bag of Words:
http://archive.ics.uci.edu/ml/machine-learning-databases/bag-of-words/readme.txt


The dataset is in a text file with the following structure:
---

docID1 wordID1 count
docID1 wordID2 count
docID1 wordID3 count
docID1 wordID4 count
...
docID2 wordID2 count
docID2 wordID5 count
docID2 wordID6 count
---

Where docIDx is an integer that identifies the document x; wordIDy is
an integer that identifies the word y ; and count is an integer with
the number of times that the wordIDy appears in the docIDx.


Example:

---

1 1 3
1 2 54
1 3 11
1 4 17
2 1 5
2 4 78
2 5 20
---

I would like to import the file into a matrix (not sparse) where:

the wordIDy would correspond to the column [,y]

the docIDx would correspond to the row [x,]

the value in [x,y] would be the count of wordIDy in the docIDx

So, for the previous example it would be like:


     [,1][,2][,3][,4][,5]

[1,]  3   54  11 17   0

[2,]  5    0   0 78  20


I don1t have a clue about how to do this.

Can someone please help me?

Thank you

Rui

        [[alternative HTML version deleted]]

______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to