Dear all

Using the randomGLM package I get an 'NA' value for the prediction of certain samples. Although it works fine for randomGLM$predictedOOB.response, it's not the case for thinGLM$predicted00B.response. Does anyone has an explanation or advice? 

Many thanks. 

===============================

> thinRGLM$predictedOOB.response

                         0            1

GSM423045.CEL 9.999997e-01 2.609077e-07

GSM423043.CEL 1.000000e+00 2.220446e-16

GSM423039.CEL 9.993901e-01 6.098978e-04

GSM423035.CEL 1.000000e+00 1.311970e-08

GSM423027.CEL           NA           NA

GSM423023.CEL 2.220446e-16 1.000000e+00

GSM423021.CEL 2.220446e-16 1.000000e+00

GSM423015.CEL 2.220446e-16 1.000000e+00

GSM423013.CEL           NA           NA

GSM423012.CEL 2.220446e-16 1.000000e+00

GSM423010.CEL 3.841227e-09 1.000000e+00

GSM364654.CEL 1.000000e+00 2.220446e-16

GSM364652.CEL 1.000000e+00 2.220446e-16

GSM364649.CEL 1.000000e+00 2.220446e-16

GSM364647.CEL 9.984750e-01 1.524975e-03

GSM364646.CEL 1.000000e+00 2.220446e-16

GSM364645.CEL 1.000000e+00 2.220446e-16

GSM364643.CEL 1.000000e+00 2.220446e-16

GSM364642.CEL           NA           NA

GSM364641.CEL 1.000000e+00 2.220446e-16

GSM364640.CEL 2.220446e-16 1.000000e+00

GSM364638.CEL 3.911849e-04 9.996088e-01

GSM364637.CEL 1.188934e-09 1.000000e+00

GSM364635.CEL 7.516210e-14 1.000000e+00

GSM364634.CEL 5.206058e-12 1.000000e+00

GSM364633.CEL 2.220446e-16 1.000000e+00