[#6781] show, format and/or print methods give error for large dimensional sparse matrices with NA's for dimnames

Date:
2022-10-04 11:58 Priority:
3

State:
Closed

Submitted by:
Luc van Schijndel (luckylluck2) Assigned to:
Nobody (None)

Hardware:
None Product:
None

Operating System:
All Component:
None

Version:
None Severity:
trivial

Resolution:
None

URL:

Summary:
show, format and/or print methods give error for large dimensional sparse matrices with NA's for dimnames

Detailed description
#throws no error with small dimensions, works fine Matrix::sparseMatrix(i = c(1,1,2), j = c(2,3,3), x = 1:3, dims = c(3,3), dimnames = list(rep(NA, 3), rep(NA, 3))) #throws an error: Matrix::sparseMatrix(i = c(1,1,2), j = c(2,3,3), x = 1:3, dims = c(100, 100), dimnames = list(rep(NA, 100), rep(NA, 100))) #OUTPUT: # 100 x 100 sparse Matrix of class "dgCMatrix" # Error in if (msg.if.not.empty && is.list(dn) && length(dn) >= 2 && # is.character(cn <- dn[[2]]) && : # missing value where TRUE/FALSE needed My guess is that the format() method throws the error, since we do get the description of the object: '100 x 100 sparse Matrix of class "dgCMatrix"'

Comments:

Message
Date: 2022-10-12 17:53 Sender: Mikael Jagan Thanks - the problem was in Matrix:::emptyColnames(), called by format(), etc. when a matrix is wide enough that column names are suppressed. Should be fixed now, with r6781. You can test by installing latest Matrix: > install.packages("Matrix", repos = "http://R-Forge.R-project.org";) You might need type = "source", as the build hasn't finished yet.

Field	Old Value	Date	By
status_id	Open	2022-10-12 17:53	jaganmn
close_date	None	2022-10-12 17:53	jaganmn