When I try to use TMultiLayerPerceptron package, it gives me a “Segmentation fault” error at trainning. Even the example mlpHiggs.C in the tutorials gives me the same error at trainning stage.
I used gdb to check where it breaks down, below is the backtrace result:
#0 0x01200242 in _int_free () from /lib/tls/libc.so.6
#1 0x0120173b in free () from /lib/tls/libc.so.6
#2 0x01152dd3 in operator delete () from /usr/lib/libstdc++.so.5
#3 0x01152e2f in operator delete[] () from /usr/lib/libstdc++.so.5
#4 0x00282ce4 in TMultiLayerPerceptron::Train ()
from /znev09a/root/lib/root/libMLP.so
#5 0x0804a68f in main ()
I remember when I was using also root 4 at fedora core 1, there was no such kind of problem. Could anybody give me a hint what’s wrong?
Thanks a lot in advance,
Zhenhai