deletions | additions       diff --git a/code/programs/Makefile b/code/programs/Makefile  index a2f3550..e46e96f 100644  --- a/code/programs/Makefile  +++ b/code/programs/Makefile 
...
DIR_EIGEN=/usr  CXX=c++  CXXOPTS=-std=c++11 -Wall -Wno-unused-variable -Wno-deprecated-declarations  TORCH=$(HOME)/torch/install/bin/th  all: bin/eigen3 r julia torch  bin/eigen3  r: R.R  R -f R.R > /dev/null  julia: julia.jl  julia julia.jl  torch: torch.lua  $(TORCH) torch.lua  bin/eigen3: eigen.cpp  $(CXX) $(CXXOPTS) -I $(DIR_EIGEN)/include/eigen3 -o eigen3 eigen.cpp  mkdir -p bin  diff --git a/code/programs/torch.lua b/code/programs/torch.lua  new file mode 100644  index 0000000..28f1a7e  --- /dev/null  +++ b/code/programs/torch.lua 
...
-- Lua test  require 'torch'  m = 5  n = 4  A = torch.ones(m, n)  B = torch.ones(n, m)  C = torch.ones(m, m)  x = torch.ones(m)  y = torch.ones(n)  z = torch.ones(n)  alpha = 1.0  beta = 1.0  ---------------------------  -- Elementary operations --  ---------------------------  -- Multiplications  ------------------  matmul = A*B  assert(matmul:nDimension() == 2)  assert(matmul:size(1) == m)  assert(matmul:size(2) == m)  matvec = A*y  assert(matvec:nDimension() == 1)  assert(matvec:size(1) == m)  -- assert(matvec:size(2) == 1)  -- error:  -- bad argument #1 to 'size' (dimension 2 out of range of 1D tensor  -- no trailing singleton dimensions  matscal = A*alpha  assert(matscal:nDimension() == 2)  assert(matscal:size(1) == m)  assert(matscal:size(2) == n)  --vecmat = x*A  --Result: error  --multiplication between 1D and 2D tensors not yet supported  vecvec = x*x  -- Result: print(vecvec:size()) - a scalar  --vecscal = x*alpha  --scalmat = alpha*A  --Result: error  --internal error in __mul: no metatable  --scalvec = alpha*x  --Result: error  --internal error in __mul: no metatable  scalscal = alpha*beta  -- Divisions  ------------  -- matmatdiv = A/A  -- Result: error  -- bad argument #2 to '?' (number expected  -- matvecdiv = A/x  -- Result: error  -- bad argument #2 to '?' (number expected  matscaldiv = A/alpha  assert(matscaldiv:nDimension() == 2)  assert(matscaldiv:size(1) == A:size(1))  assert(matscaldiv:size(2) == A:size(2))  -- vecmatdiv = x/A  -- Result: error  -- bad argument #2 to '?' (number expected  --vecvecdiv = x/x  -- Result: error  -- bad argument #2 to '?' (number expected  vecscaldiv = x/alpha  assert(vecscaldiv:nDimension() == 1)  assert(vecscaldiv:size(1) == x:size(1))  -- scalmatdiv = alpha/A  -- Result: error  -- internal error in __div: no metatable  -- scalvecdiv = alpha/x  -- Result: error  -- internal error in __div: no metatable  scalscaldiv = alpha/beta  -- Result: scalar  -- Transpositions  -----------------  mattrans = A:t()  assert(mattrans:nDimension() == 2)  assert(mattrans:size(1) == A:size(2))  assert(mattrans:size(2) == A:size(1))  --vectrans = x:t()  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  --scaltrans = alpha:t()  --Result: error  -- attempt to index global 'alpha' (a number value)  -----------------  -- Expressions --  -----------------  --inner = x:t()*x  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  inner = x:dot(x)  --outer = x*y:t()  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  outer = torch.ger(x, y)  --quadratic = x:t()*C*x  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  quadratic = x:dot(C*x)  --bilinear = x:t()*A*y  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  bilinear = x:dot(A*y)  --rayleigh = x:t()*C*x/(x:t()*x)  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  rayleigh = x:dot(C*x)/(x:dot(x))  ----------------  -- Identities --  ----------------  --assert( (x:t()):t() == x)  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  --assert( (A*y):t() == y:t() * A:t() )  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)  --assert( (x*y:t())*z == x*(y:t()*z) )  --Result: error  --calling 't' on bad self (Tensor must have 2 dimensions)