[Tutor] PyBrain: Question on creating customized DataSet

[prudhvi nethi]

Hi,

I'm trying to create a data set for using it with BackpropTrainer
*(PyBrain)*

The model is that i have a bunch of inputs (which can have multiple
values). Resulting in some Output Kx( different values).

Sample data ( A subset of the data )

 Input (x)

Input (y)

Input (z)

Input(a)

Output(k)

X1

Y1

Z1

A1

K1

X1

Y2

Z2

A3

K2

X2

Y3

Z2

A2

K1





Where Input x can have any of the following value in a specific entry:
(X1,X2,x2)



*An example from PyBrain Documentation for a simple XOR data set modeling:*


>>> ds.addSample((0, 0), (0,))>>> ds.addSample((0, 1), (1,))>>> ds.addSample((1, 0), (1,))>>> ds.addSample((1, 1), (0,))


As the function is simple here (XOR) and the inputs and o/p's can be
represented in binary format. How can i model a data set for a slightly
complex like mine ?

*Please let me know if i mis-speak or interpret something here.*

-Prudhvi