func init() {
mat64.Register(cblas.Blas{})
// Set up all of the test sinks
for i, initer := range sinkIniters {
s := NewTrainer(initer.inputDim, initer.outputDim, initer.nFeatures, initer.kernel)
s.features = randomMat(initer.nFeatures, initer.inputDim)
s.featureWeights = randomMat(initer.nFeatures, initer.outputDim)
s.b = randomSlice(initer.nFeatures)
testSinks = append(testSinks, s)
if initer.name == "" {
initer.name = strconv.Itoa(i)
}
}
}
// Activate propagates the given input matrix (with) across the network
// a certain number of times (up to maxIterations).
//
// The with matrix should be size * size elements, with only the values
// of input neurons set (everything else should be zero).
//
// If the network is conceptually organised into layers, maxIterations
// should be set to the number of layers.
//
// This function overwrites whatever's stored in its first argument.
func (n *Network) Activate(with *mat64.Dense, maxIterations int) {
// Add bias and feed to activation
biasFunc := func(r, c int, v float64) float64 {
return v + n.biases[r]
}
activFunc := func(r, c int, v float64) float64 {
return n.funcs[r].Forward(v)
}
tmp := new(mat64.Dense)
tmp.Clone(with)
mat64.Register(cblas.Blas{})
// Main loop
for i := 0; i < maxIterations; i++ {
with.Mul(n.weights, with)
with.Apply(biasFunc, with)
with.Apply(activFunc, with)
}
}
func init() {
mat64.Register(goblas.Blas{})
}
func init() {
mat64.Register(goblas.Blas{}) // use a go-based blas library
dbw.Register(goblas.Blas{})
}
