Re: [R] unique possible bug

2011-10-06 Thread Patrick McCann
The error I am referring to is in unique.c in Base R, it cannot accomodate greater than 2^29 values, even though it appears the overflow protection should be 2^30. The only relevance of the arules package is I was using it while I discovered this issue. Thanks, Patrick 2011/10/6 Uwe Ligges : >

Re: [R] unique possible bug

2011-10-06 Thread Uwe Ligges
I see. For now: Yes, you need to change and recompile. I will take a look what was actually changed and will run some test cases. Best, Uwe On 06.10.2011 16:50, Patrick McCann wrote: The error I am referring to is in unique.c in Base R, it cannot accomodate greater than 2^29 values, even thou

Re: [R] unique possible bug

2011-10-06 Thread Uwe Ligges
On 05.10.2011 22:15, Patrick McCann wrote: Hi, I am trying to read in a rather large list of transactions using the arules library. You mean the arules package? It seems in the coerce method into the dgCmatrix, it somewhere calls unique. Unique.c throws an error when n> 536870912; howev

[R] unique possible bug

2011-10-05 Thread Patrick McCann
Hi, I am trying to read in a rather large list of transactions using the arules library. It seems in the coerce method into the dgCmatrix, it somewhere calls unique. Unique.c throws an error when n > 536870912; however, when 4*n was modified to 2*n in 2004, the overflow protection should have cha