[
https://issues.apache.org/jira/browse/SYSTEMML-294?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15282812#comment-15282812
]
Mike Dusenberry commented on SYSTEMML-294:
------------------------------------------
Merged in [commit ea49e62 |
https://github.com/apache/incubator-systemml/commit/ea49e620ce846c41c8b9bb9b729986a9aa7fd4d5].
> Print matrix capability
> -----------------------
>
> Key: SYSTEMML-294
> URL: https://issues.apache.org/jira/browse/SYSTEMML-294
> Project: SystemML
> Issue Type: Improvement
> Components: APIs
> Reporter: Deron Eriksson
> Assignee: Nakul Jindal
>
> The ability to output a matrix or a part of a matrix to standard output would
> be very useful to have while developing DML and PyDML scripts. By having the
> functionality as a built-in function, capabilities such as alignment of cell
> values could be performed. Such a built-in function could also optionally
> take parameters to limit the output to a certain subset of cells in the
> matrix. Such a function would be useful for small matrices. An error could be
> thrown if an attempt is made to use the function with a matrix that is too
> large.
> Here's an example of a printMatrix function in DML:
> {code}
> printMatrix = function(matrix[double] mat) {
> for (i in 1:nrow(mat)) {
> colVals = '| '
> for (j in 1:ncol(mat)) {
> n = mat[i,j]
> colVals = colVals + as.scalar(n) + ' | '
> }
> print(colVals)
> }
> }
> m = matrix("1 2 3 4 5 6 7 8 9 10 11 12", rows=4, cols=3)
> z = printMatrix(m)
> {code}
> Here's an example of a printMatrix function in PyDML:
> {code}
> def printMatrix(mat:matrix[float]):
> for (i in 1:nrow(mat)):
> colVals = '| '
> for (j in 1:ncol(mat)):
> n = mat[i,j]
> colVals = colVals + scalar(n) + ' | '
> print(colVals)
> m = full("1 2 3 4 5 6 7 8 9 10 11 12", rows=4, cols=3)
> z = printMatrix(m)
> {code}
> Here is the output:
> {code}
> | 1.0 | 2.0 | 3.0 |
> | 4.0 | 5.0 | 6.0 |
> | 7.0 | 8.0 | 9.0 |
> | 10.0 | 11.0 | 12.0 |
> {code}
> Notice how alignment readily becomes an issue if an attempt is made to do
> this with DML or PyDML.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
