Light Cone Reconstruction of States  Predictive State...
This function merges states i_1, …, i_j into a new, single state i_1 by adding corresponding columns of the weight matrix (\mathbf{W}_{i_1} = \mathbf{W}_{i_1} + … + \mathbf{W}_{i_j}) and removing columns i_2, …, i_j.
1  merge_states(states, weight.matrix)

states 
vector of length 1 ≤q j ≤q K with the states i_1, …, i_j \subset \lbrace 1, …, K \rbrace that should be merged; no repeating state labels allowed. 
weight.matrix 
N \times K weight matrix 
1 2 3 4 5 6 7 8 9 10 11 12 13  set.seed(10)
WW < matrix(c(rexp(1000, 1/10), runif(1000)), ncol = 5, byrow = FALSE)
WW < normalize(WW)
image2(WW, density = TRUE)
## Not run:
merge_states(c(1, 1, 5), WW) # error since states were repeated
## End(Not run)
WW_new < merge_states(c(1, 3, 5), WW)
par(mfrow = c(1, 2), mar = c(1, 1, 2, 1))
image2(WW, main = paste(ncol(WW), "states"), legend = FALSE)
image2(WW_new, main = paste(ncol(WW_new), "states"), legend = FALSE)

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.
Please suggest features or report bugs with the GitHub issue tracker.
All documentation is copyright its authors; we didn't write any of that.