nn.Recurrent is unit tested

1 files changed, 18 insertions, 11 deletions

diff --git a/test/test-all.lua b/test/test-all.lua
index 0d666e0..10f9166 100644
--- a/test/test-all.lua
+++ b/test/test-all.lua
@@ -307,7 +307,7 @@ function nnxtest.Recurrent()
    local inputSize = 10
    local hiddenSize = 12
    local outputSize = 7
-   local nSteps = 5   
+   local nSteps = 5
    local inputModule = nn.Linear(inputSize, outputSize)
    local transferModule = nn.Sigmoid()
    -- test MLP feedback Module (because of Module:representations())
@@ -316,12 +316,13 @@ function nnxtest.Recurrent()
    feedbackModule:add(nn.Sigmoid())
    feedbackModule:add(nn.Linear(hiddenSize, outputSize))
    local mlp = nn.Recurrent(outputSize, inputModule, feedbackModule, transferModule:clone())
-   inputModule = mlp.inputModule:clone()
-   feedbackModule = mlp.feedbackModule:clone()
    
    local gradOutputs, outputs = {}, {}
+   -- inputs = {inputN, {inputN-1, {inputN-2, ...}}}}}
    local inputs
    local startModule = mlp.startModule:clone()
+   inputModule = mlp.inputModule:clone()
+   feedbackModule = mlp.feedbackModule:clone()
    
    mlp:zeroGradParameters()
    for step=1,nSteps do
@@ -348,15 +349,14 @@ function nnxtest.Recurrent()
       end
    end
    -- backward propagate through time (BPTT)
-   mlp:backwardThroughTime()
+   local gradInput = mlp:backwardThroughTime()
    
-   -- input = {inputN, {inputN-1, {inputN-2, ...}}}}}
    local mlp2
-   local outputs2 = {}
+   local outputModules = {}
    for step=1,nSteps do
       local inputModule_ = inputModule:clone()
       local outputModule = transferModule:clone()
-      table.insert(outputs2, outputModule.output)
+      table.insert(outputModules, outputModule)
       inputModule_:share(inputModule, 'weight', 'gradWeight', 'bias', 'gradBias')
       if step == 1 then
          local initialModule = nn.Sequential()
@@ -384,21 +384,28 @@ function nnxtest.Recurrent()
    local output2 = mlp2:forward(inputs)
    mlp2:zeroGradParameters()
    local gradInput2 = mlp2:backward(inputs, gradOutputs[#gradOutputs])
+   for step=1,nSteps-1 do
+      gradInput2 = gradInput2[2]
+   end   
    
+   mytester:assertTensorEq(gradInput, gradInput2, 0.000001, "recurrent gradInput")
    mytester:assertTensorEq(outputs[#outputs], output2, 0.000001, "recurrent output")
    for step=1,nSteps do
-      local output, output2 = outputs[step], outputs2[step]
-      mytester:assertTensorEq(output, output2, 0.000001, "recurrent output step="..step)
+      local output, outputModule = outputs[step], outputModules[step]
+      mytester:assertTensorEq(output, outputModule.output, 0.000001, "recurrent output step="..step)
    end
    
    local mlp3 = nn.Sequential()
-   mlp3:add(startModule):add(inputModule):add(recurrentModule)
+   mlp3:add(startModule):add(inputModule):add(feedbackModule)
    local params2, gradParams2 = mlp3:parameters()
    local params, gradParams = mlp:parameters()
    mytester:assert(#params2 == #params, 'missing parameters')
    mytester:assert(#gradParams == #params, 'missing gradParameters')
    for i=1,#params do
-      mytester:assertTensorEq(gradParams[i], gradParams2[i], 0.000001, 'gradParameter error ' .. 1)
+      if i > 1 then
+         gradParams2[i]:div(nSteps)
+      end
+      mytester:assertTensorEq(gradParams[i], gradParams2[i], 0.000001, 'gradParameter error ' .. i)
    end
 end