SoftMaxTree:forward works (debugged, unit tested)

1 files changed, 17 insertions, 4 deletions

diff --git a/SoftMaxTree.lua b/SoftMaxTree.lua
index f87a35e..e8aa21c 100644
--- a/SoftMaxTree.lua
+++ b/SoftMaxTree.lua
@@ -10,6 +10,7 @@ local SoftMaxTree, parent = torch.class('nn.SoftMaxTree', 'nn.Module')
 -- differ setup after init
 -- verify that each parent has a parent (except root)
 -- nodeIds - 1?
+-- types
 ------------------------------------------------------------------------
 
 function SoftMaxTree:__init(inputSize, hierarchy, rootId, verbose)
@@ -49,6 +50,7 @@ function SoftMaxTree:__init(inputSize, hierarchy, rootId, verbose)
          print("in order to waste less memory on indexes.")
       end
    end
+   
    self.nChildNode = nChildNode
    self.nParentNode = nParentNode
    self.minNodeId = minNodeId
@@ -73,7 +75,7 @@ function SoftMaxTree:__init(inputSize, hierarchy, rootId, verbose)
    for parentId, children in pairs(hierarchy) do
       local node = self.parentChildren:select(1, parentId)
       node[1] = start
-      local nChildren = children:size()
+      local nChildren = children:size(1)
       node[2] = nChildren
       self.childIds:narrow(1, start, nChildren):copy(children)
       start = start + nChildren
@@ -81,7 +83,7 @@ function SoftMaxTree:__init(inputSize, hierarchy, rootId, verbose)
    
    -- index of parent by childId
    self.childParent = torch.IntTensor(self.maxChildId, 2):fill(-1)
-   for parendIdx=1,self.parentIds:size(1) do
+   for parentIdx=1,self.parentIds:size(1) do
       local parentId = self.parentIds[parentIdx]
       local node = self.parentChildren:select(1, parentId)
       local start = node[1]
@@ -99,12 +101,23 @@ function SoftMaxTree:__init(inputSize, hierarchy, rootId, verbose)
    self._linearOutput = torch.Tensor()
    self._logSoftMaxOutput = torch.Tensor()
    self._narrowOutput = torch.Tensor()
-   -- used to store pointers to intermediate outputs
-   --self._nodes = torch.Tensor()
    
+   print("parentIds", self.parentIds)
+   print("parentChildren", self.parentChildren)
+   print("children", self.childIds)
+   print("childParent", self.childParent)
    self:reset()
 end
 
+function SoftMaxTree:getNodeParameters(parentId)
+   local node = self.parentChildren:select(1,parentId)
+   local start = node[1]
+   local nChildren = node[2]
+   local weight = self.weight:narrow(1, start, nChildren)
+   local bias = self.bias:narrow(1, start, nChildren)
+   return weight, bias
+end
+
 function SoftMaxTree:reset(stdv)
    if stdv then
       stdv = stdv * math.sqrt(3)