Classification process do not work

What steps will reproduce the problem?

1.

2.Create & train dataset (see outpout)
Summary of 'dataset/proj/dataset_proj.fit' (120 samples total, 1 samples per 
image):
'Class label' (interpreted value) number of samples.
'0005'  (5) 10
'0010'  (10)    10
'0011'  (11)    10
'0014'  (14)    10
'0025'  (25)    10
'0038'  (38)    10
'0040'  (40)    10
'0046'  (46)    10
'0050'  (50)    10
'0064'  (64)    10
'0075'  (75)    10
'0093'  (93)    10
Class labels are purely numeric

ARGS : train params : ld50r0.5i10S800B0,0,800,350
     : test params : ld50S800B0,0,800,350i6f1

3. wndchrm classify -ld50S800B0,0,800,350f1 dataset/proj/dataset_proj.fit 
/root/conv-3.TIF 

What is the expected output? What do you see instead?
Expected ouput  : a document classification 

See instead:
----------
image   norm. 
fact.   p(0005) p(0010) p(0011) p(0014) p(0025) p(0038) p(0040) p(0046) p(0050) p(
0064)   p(0075) p(0093) act. class  pred. class pred. val.
/root/conv-9-0064.TIF   1.61e-27    0.027   0.973   0.000   0.000   0.000   0.000   0.000   0.000   0
.000    0.000   0.000   0.000   *   0010    9.871

...
----------
WARNING: Test set class label '' does not match any training set class.  Marked 
with '*'.: Numerical argument out of domain

What version of the product are you using? On what operating system?
 With both version wndchrm-1.32b.309 & wndchrm-1.50.727
On debian squeeze Linux Basesys 2.6.32-5-amd64 #1 SMP Fri May 10 08:43:19 UTC 
2013 x86_64 GNU/Linux

Please provide any additional information below.

All the 3 documents tested come from the training dataset . here is the dataset 
test result :
 Accuracy: 0.88 of total (P=6.3e-96) 
0.88 ± 0.57 Avg per Class Correct of total
Pearson correlation coefficient: 0.94 (P=4.38e-57) 
Mean absolute difference: 4.0444 
Full details
Total    Total tested: 120
Total correct: 105
Accuracy: 87.5% of total (P=6.3e-96)
Classification accuracy: 87.5 +/- 5.9% (95% confidence, normal approx 
confidence interval)
Pearson correlation coefficient: 0.94 (avg P=4.38e-57) 
Mean absolute difference: 4.0444
Original issue reported on code.google.com by cner...@gmail.com on 27 Nov 2013 at 9:08
kzwkt / wnd-charm

Classification process do not work #66
