MixtureTable unit tests

1 files changed, 21 insertions, 0 deletions

diff --git a/test/test.lua b/test/test.lua
index 1eb571e..8eab9a1 100644
--- a/test/test.lua
+++ b/test/test.lua
@@ -1954,6 +1954,27 @@ function nntest.SelectTable()
    equal(gradInput[nonIdx[idx]], zeros[nonIdx[idx]], "gradInput[nonIdx] dimension " .. idx)
 end
 
+function nntest.MixtureTable()
+   local expertInput = torch.randn(5,3,6)
+   local input = {
+      torch.rand(5,3), 
+      {expertInput:select(2,1), expertInput:select(2,2), expertInput:select(2,3)}
+   }
+   local gradOutput = torch.randn(5,6)
+   local module = nn.MixtureTable()
+   local output = module:forward(input)
+   local output2 = torch.cmul(input[1]:view(5,3,1):expand(5,3,6), expertInput):sum(2)
+   mytester:assertTensorEq(output, output2, 0.000001, "mixture output")
+   local gradInput = module:backward(input, gradOutput)
+   local gradOutput2 = torch.view(gradOutput, 5, 1, 6):expandAs(expertInput)
+   local gaterGradInput2 = torch.cmul(gradOutput2, expertInput):sum(3):select(3,1)
+   mytester:assertTensorEq(gradInput[1], gaterGradInput2, 0.000001, "mixture gater gradInput")
+   local expertGradInput2 = torch.cmul(input[1]:view(5,3,1):expand(5,3,6), gradOutput:view(5,1,6):expand(5,3,6))
+   for i, expertGradInput in ipairs(gradInput[2]) do
+      mytester:assertTensorEq(expertGradInput, expertGradInput2:select(2,i), 0.000001, "mixture expert "..i.." gradInput")
+   end
+end
+
 function nntest.View()
    local input = torch.rand(10)
    local template = torch.rand(5,2)