#6 SEGV in StumpLearner::classify

[Stephen T. Pope]

Hi all,

I'm new to using Multiboost, and am getting repeatable seg-faults on my data. Does anyone have any experience with this?

Here's the stack:

(gdb) run -train FMAK_arff2.csv 10

Starting program: /Users/stp/Code/MIR/MultiBoost/multiboost -train FMAK_arff2.csv 10

Loading file FMAK_arff2.csv...Done!
Sorting data...Done!
Learning in progress...

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_PROTECTION_FAILURE at address: 0x00000000
0x0002784e in MultiBoost::StumpLearner::classify ()

(gdb) where

#0 0x0002784e in MultiBoost::StumpLearner::classify ()
#1 0x00001e60 in MultiBoost::AdaBoostMHLearner::updateWeights ()
#2 0x0000312e in MultiBoost::AdaBoostMHLearner::run ()
#3 0x0002c1a3 in main ()

(gdb)

Here's a typical line from the data file (15000 records, many features):

Rock,0.833341,0.790646,0.864314,0.286719,0.191326,0.100672,0.0546172,0.0303422,0.0270417,0.492952,0,0.0759092,0.0715664,0.078329,0.0682285,0.255632,0.01413,0.0140967,0.0140399,0.0139591,0.0138548,0.0682285,0.0682285,0.434986,0.0298234,0,0.9,0.0706565,0.499994,0.00101918,0.515099,0.232283,0.945134,0,0,0,0,0,0,0,0,0.151095,0.0927906,0.281136,0.290368,0.192618,0.198883,4.08E-05,1.14E-07,0.00509301,0.0475029,2.05E-07,0.00509301,0.163385,0,0.0067639,0.00570942,0.0134778,0.00554504,0.00554504,0.00554504,0.227835,0.00892062,0.00890013,0.00886574,0.00881777,0.00875609

Does this look kosher?

...any reply appreciated...

stp