I am using LIBSVM and when I give a command for scale up my test data has an saying error saying 'segmentation fault'. What does it mean and how can I remove it? My data is too large.
Hello,
Segmentation faults usually suggest bugged code and/or flawed compiler. What you can actually do is:
1. Upgrade you ubuntu. This is necessary since modern versions of gcc may not be compatible with your (11.04) kernel and this is a know source of problems;
2. Update your gcc. This may solve your problem. Usually done through
sudo apt-get update && sudo apt-get upgrade
3. After that, recompile everything. Also check if your compiler flags are safe; you don't want to risk building broken files.
4. then try again
PS. Another common source of seg faults is the system trying to access virtual memmory that doesn't exist. Maybe your software/library has some sort of memmory cap configuration. If that's the case, it's worth using it.
Hope this helps.
All the best,
Luiz.
Same suggestion: Try upgrading to Ubuntu 12.04.2 LTS (Precise Pangolin)
Since you mentioned large data in your post: Try with less data. It could be some kind of buffer overflow caused by excess data not being handled properly by the program itself. If so, you need to look for ways to increase the tolerance of the program (configuration, command-line parameters, recompiling after making changes to the code).
Metej Lexa, I tried with less data, but still this error has came, i tried in many system having good configuration of i3 or i5 processer and 3 -4 gb ram.
ok... Segmentation fault report usually come for insufficient memory. Try with higher RAM. Thats the solution. is it a sequence data? if so. if your fastQ file has more than 3 lakhs reads (usually) it will show segmentation nfault with 4 gb RAM.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools