negative tpr,fpr

[cycy1013]

when i run simulateModel.run() sometimes tpr,fpr could be negative?
tpr,fpr belongs to [0,1], why negative could emerge?

[ishanuc]

what data does this happen with?

[cycy1013]

https://github.com/cycy1013/Cynet-master1/tree/master/stock2015_2017
i use Cynet to predict stock close price

all in stock2015_2017 directory

the method is treating trade data as spacial factor

but now the result is confusing

for example:in 3modeluse120models#ALL#close.csv

tail content:

close,714,1,0.690375,0.309625,1,ALL,-1.0

close,715,1,0.676619,0.323381,1,ALL,-1.0

close,716,1,0.666565,0.333435,1,ALL,-1.0

close,717,0,0.6784,0.3216,1,ALL,-1.0

close,718,0,0.672881,0.327119,1,ALL,-1.0

in 3model_120_1_0.res content:

./stock2015_2017/models/3model,ALL,close,120,0.666663,0.333333,-1.0,1

1.why tpr,fpr is 0.333333,-1.0 ,come from where ?

2.why fpr is negative?

3.haven't seen valuable result in predicting stock close price?