'memory not mapped'

[onmikula]

Hi,
I wanted to use bi.clique function on a bit larger data set (matrix of dimension 12481 x 24:
mat_12481x24.txt), but got the following error:

*** caught segfault ***
address 0x111ea6000, cause 'memory not mapped'

The function worked well with fewer columns (matrix of dimension 12481 x 10:
mat_12481x10.txt
), but not with fewer rows (5200 x 24:
mat_5200x24.txt
), although the number of entries was approximately the same in both cases.

Would you be so kind and check what is the problem?
Thank you in advance!

Ondrej

[dpuelz]

Hey, I have a similar issue too, just noted on issue #6 .. hopefully we can get it figured out!

David

[YupingLu]

Hi all, I'll check the problem later this weekend.

[dpuelz]

Hi all, I'll check the problem later this weekend.

Hey Yuping, thanks for the help! Just curious, have you made any progress on a solution?

[onmikula]

Hello Yuping,
do you thing the problem is going to be fixed any time soon?
Ondrej

[YupingLu]

Hi Ondrej,

Sorry about the delay. I'm busy with multiple things recently. I'll look at it again at the end of this month. Hope the problem will be fixed soon.

Yuping

[onmikula]

Thank you for your reply. If the problem was fixed, I would appreciate it hugely. It would allow me to use the algorithm where it is possibly most useful (i.e. in graphs with large number of vertices of either type).
Best wishes,
Ondrej

[onmikula]

Hi Yuping,
are there any news? I have found yet another nice application for the maximal biclique enumeration, so you can see your package is really needed.
Best,
Ondrej

[dpuelz]

Hey Yuping, Agreed. Let us know if there’s a fix. Thanks! — David On Feb 15, 2019, at 12:07 PM, onmikula <notifications@github.com<mailto:notifications@github.com>> wrote: Hi Yuping, are there any news? I have found yet another nice application for the maximal biclique enumeration, so you can see your package is really needed. Best, Ondrej — You are receiving this because you commented. Reply to this email directly, view it on GitHub<#7 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AJffJDTxk5-x-XnScn_1E9iuFmt9LyH8ks5vNulGgaJpZM4Zkyuk>.

[YupingLu]

Hi Ondrej,

I finally got back to this. R doesn't allow C++ to ouput bicliques to screen. If the number of bicliques is big, R will not have enough memory to store all bicliques. That's why you get this error. I added a new parameter to only get the statistics without storing bicliques. You can get the top bicliques by setting lleast and rleast. If you still want to generate all the bicliques, we also have a C++ version available.

# Only get the statistics
library(biclique)
test=bi.clique('mat_5200x24.txt', getclique=0, filetype=1)

Number of left vertices : 5200
Number of right vertices : 24
Number of edges : 98874
Number of bicliques : 16731423
Maximum edge biclique : K2335,12
Maximum vertex biclique : K4898,1

# Get the maximum edge biclique
test=bi.clique('mat_5200x24.txt', lleast = 2335, rleast = 12, filetype=1)

Yuping