Resolve merge.

2026-04-10 11:34:33 +08:00 · 2016-06-23 15:08:03 -07:00
parent f1f2ff8208 361dbd246d
commit d39df320d2
121 changed files with 3058 additions and 1132 deletions
--- a/unsupported/test/CMakeLists.txt
+++ b/unsupported/test/CMakeLists.txt
@@ -112,6 +112,10 @@ ei_add_test(kronecker_product)

 # TODO: The following test names are prefixed with the cxx11 string, since historically
 # the tests depended on c++11. This isn't the case anymore so we ought to rename them.
+# FIXME: Old versions of MSVC fail to compile this code, so we just disable these tests
+# when using visual studio. We should make the check more strict to enable the tests for
+# newer versions of MSVC.
+if (NOT CMAKE_CXX_COMPILER_ID STREQUAL "MSVC")
 ei_add_test(cxx11_float16)
 ei_add_test(cxx11_tensor_dimension)
 ei_add_test(cxx11_tensor_map)
@@ -130,7 +134,8 @@ ei_add_test(cxx11_tensor_io)
 if("${CMAKE_SIZEOF_VOID_P}" EQUAL "8")
  # This test requires __uint128_t which is only available on 64bit systems 
  ei_add_test(cxx11_tensor_uint128)
-endif() 
+endif()
+endif()

 if(EIGEN_TEST_CXX11)
  # It should be safe to always run these tests as there is some fallback code for
--- a/unsupported/test/cxx11_tensor_cuda.cu
+++ b/unsupported/test/cxx11_tensor_cuda.cu
@@ -1019,6 +1019,153 @@ void test_cuda_erfc(const Scalar stddev)
  cudaFree(d_out);
 }

+template <typename Scalar>
+void test_cuda_betainc()
+{
+  Tensor<Scalar, 1> in_x(125);
+  Tensor<Scalar, 1> in_a(125);
+  Tensor<Scalar, 1> in_b(125);
+  Tensor<Scalar, 1> out(125);
+  Tensor<Scalar, 1> expected_out(125);
+  out.setZero();
+
+  Scalar nan = std::numeric_limits<Scalar>::quiet_NaN();
+
+  Array<Scalar, 1, Dynamic> x(125);
+  Array<Scalar, 1, Dynamic> a(125);
+  Array<Scalar, 1, Dynamic> b(125);
+  Array<Scalar, 1, Dynamic> v(125);
+
+  a << 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
+      0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999,
+      0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999,
+      0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 0.999, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379, 999.999, 999.999,
+      999.999, 999.999, 999.999, 999.999, 999.999, 999.999, 999.999, 999.999,
+      999.999, 999.999, 999.999, 999.999, 999.999, 999.999, 999.999, 999.999,
+      999.999, 999.999, 999.999, 999.999, 999.999, 999.999, 999.999;
+
+  b << 0.0, 0.0, 0.0, 0.0, 0.0, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379, 0.999,
+      0.999, 0.999, 0.999, 0.999, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379, 999.999, 999.999,
+      999.999, 999.999, 999.999, 0.0, 0.0, 0.0, 0.0, 0.0, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.999, 0.999, 0.999, 0.999, 0.999, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 999.999, 999.999, 999.999, 999.999, 999.999, 0.0, 0.0,
+      0.0, 0.0, 0.0, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379, 0.999,
+      0.999, 0.999, 0.999, 0.999, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379, 999.999, 999.999,
+      999.999, 999.999, 999.999, 0.0, 0.0, 0.0, 0.0, 0.0, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.999, 0.999, 0.999, 0.999, 0.999, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 999.999, 999.999, 999.999, 999.999, 999.999, 0.0, 0.0,
+      0.0, 0.0, 0.0, 0.03062277660168379, 0.03062277660168379,
+      0.03062277660168379, 0.03062277660168379, 0.03062277660168379, 0.999,
+      0.999, 0.999, 0.999, 0.999, 31.62177660168379, 31.62177660168379,
+      31.62177660168379, 31.62177660168379, 31.62177660168379, 999.999, 999.999,
+      999.999, 999.999, 999.999;
+
+  x << -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8,
+      1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5,
+      0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2,
+      0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1,
+      0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1,
+      -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8,
+      1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5,
+      0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2,
+      0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1, -0.1, 0.2, 0.5, 0.8, 1.1;
+
+  v << nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan,
+      nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan, nan,
+      nan, nan, 0.47972119876364683, 0.5, 0.5202788012363533, nan, nan,
+      0.9518683957740043, 0.9789663010413743, 0.9931729188073435, nan, nan,
+      0.999995949033062, 0.9999999999993698, 0.9999999999999999, nan, nan,
+      0.9999999999999999, 0.9999999999999999, 0.9999999999999999, nan, nan, nan,
+      nan, nan, nan, nan, 0.006827081192655869, 0.0210336989586256,
+      0.04813160422599567, nan, nan, 0.20014344256217678, 0.5000000000000001,
+      0.7998565574378232, nan, nan, 0.9991401428435834, 0.999999999698403,
+      0.9999999999999999, nan, nan, 0.9999999999999999, 0.9999999999999999,
+      0.9999999999999999, nan, nan, nan, nan, nan, nan, nan,
+      1.0646600232370887e-25, 6.301722877826246e-13, 4.050966937974938e-06, nan,
+      nan, 7.864342668429763e-23, 3.015969667594166e-10, 0.0008598571564165444,
+      nan, nan, 6.031987710123844e-08, 0.5000000000000007, 0.9999999396801229,
+      nan, nan, 0.9999999999999999, 0.9999999999999999, 0.9999999999999999, nan,
+      nan, nan, nan, nan, nan, nan, 0.0, 7.029920380986636e-306,
+      2.2450728208591345e-101, nan, nan, 0.0, 9.275871147869727e-302,
+      1.2232913026152827e-97, nan, nan, 0.0, 3.0891393081932924e-252,
+      2.9303043666183996e-60, nan, nan, 2.248913486879199e-196,
+      0.5000000000004947, 0.9999999999999999, nan;
+
+  for (int i = 0; i < 125; ++i) {
+    in_x(i) = x(i);
+    in_a(i) = a(i);
+    in_b(i) = b(i);
+    expected_out(i) = v(i);
+  }
+
+  std::size_t bytes = in_x.size() * sizeof(Scalar);
+
+  Scalar* d_in_x;
+  Scalar* d_in_a;
+  Scalar* d_in_b;
+  Scalar* d_out;
+  cudaMalloc((void**)(&d_in_x), bytes);
+  cudaMalloc((void**)(&d_in_a), bytes);
+  cudaMalloc((void**)(&d_in_b), bytes);
+  cudaMalloc((void**)(&d_out), bytes);
+
+  cudaMemcpy(d_in_x, in_x.data(), bytes, cudaMemcpyHostToDevice);
+  cudaMemcpy(d_in_a, in_a.data(), bytes, cudaMemcpyHostToDevice);
+  cudaMemcpy(d_in_b, in_b.data(), bytes, cudaMemcpyHostToDevice);
+
+  Eigen::CudaStreamDevice stream;
+  Eigen::GpuDevice gpu_device(&stream);
+
+  Eigen::TensorMap<Eigen::Tensor<Scalar, 1> > gpu_in_x(d_in_x, 125);
+  Eigen::TensorMap<Eigen::Tensor<Scalar, 1> > gpu_in_a(d_in_a, 125);
+  Eigen::TensorMap<Eigen::Tensor<Scalar, 1> > gpu_in_b(d_in_b, 125);
+  Eigen::TensorMap<Eigen::Tensor<Scalar, 1> > gpu_out(d_out, 125);
+
+  gpu_out.device(gpu_device) = betainc(gpu_in_a, gpu_in_b, gpu_in_x);
+
+  assert(cudaMemcpyAsync(out.data(), d_out, bytes, cudaMemcpyDeviceToHost, gpu_device.stream()) == cudaSuccess);
+  assert(cudaStreamSynchronize(gpu_device.stream()) == cudaSuccess);
+
+  for (int i = 1; i < 125; ++i) {
+    if ((std::isnan)(expected_out(i))) {
+      VERIFY((std::isnan)(out(i)));
+    } else {
+      VERIFY_IS_APPROX(out(i), expected_out(i));
+    }
+  }
+
+  cudaFree(d_in_x);
+  cudaFree(d_in_a);
+  cudaFree(d_in_b);
+  cudaFree(d_out);
+}
+
+
 void test_cxx11_tensor_cuda()
 {
  CALL_SUBTEST_1(test_cuda_elementwise_small());
@@ -1086,5 +1233,8 @@ void test_cxx11_tensor_cuda()

  CALL_SUBTEST_5(test_cuda_igamma<double>());
  CALL_SUBTEST_5(test_cuda_igammac<double>());
+
+  CALL_SUBTEST_6(test_cuda_betainc<float>());
+  CALL_SUBTEST_6(test_cuda_betainc<double>());
 #endif
 }
--- a/unsupported/test/cxx11_tensor_io.cpp
+++ b/unsupported/test/cxx11_tensor_io.cpp
@@ -13,6 +13,20 @@
 #include <Eigen/CXX11/Tensor>


+template<int DataLayout>
+static void test_output_0d()
+{
+  Tensor<int, 0, DataLayout> tensor;
+  tensor() = 123;
+
+  std::stringstream os;
+  os << tensor;
+
+  std::string expected("123");
+  VERIFY_IS_EQUAL(std::string(os.str()), expected);
+}
+
+
 template<int DataLayout>
 static void test_output_1d()
 {
@@ -26,6 +40,12 @@ static void test_output_1d()

  std::string expected("0\n1\n2\n3\n4");
  VERIFY_IS_EQUAL(std::string(os.str()), expected);
+
+  Eigen::Tensor<double,1,DataLayout> empty_tensor(0);
+  std::stringstream empty_os;
+  empty_os << empty_tensor;
+  std::string empty_string;
+  VERIFY_IS_EQUAL(std::string(empty_os.str()), empty_string);
 }


@@ -101,6 +121,8 @@ static void test_output_const()

 void test_cxx11_tensor_io()
 {
+  CALL_SUBTEST(test_output_0d<ColMajor>());
+  CALL_SUBTEST(test_output_0d<RowMajor>());
  CALL_SUBTEST(test_output_1d<ColMajor>());
  CALL_SUBTEST(test_output_1d<RowMajor>());
  CALL_SUBTEST(test_output_2d<ColMajor>());
--- a/unsupported/test/cxx11_tensor_scan.cpp
+++ b/unsupported/test/cxx11_tensor_scan.cpp
@@ -38,6 +38,30 @@ static void test_1d_scan()
    }
 }

+template <int DataLayout, typename Type=float>
+static void test_1d_inclusive_scan()
+{
+    int size = 50;
+    Tensor<Type, 1, DataLayout> tensor(size);
+    tensor.setRandom();
+    Tensor<Type, 1, DataLayout> result = tensor.cumsum(0, true);
+
+    VERIFY_IS_EQUAL(tensor.dimension(0), result.dimension(0));
+
+    float accum = 0;
+    for (int i = 0; i < size; i++) {
+      VERIFY_IS_EQUAL(result(i), accum);
+      accum += tensor(i);
+    }
+
+    accum = 1;
+    result = tensor.cumprod(0, true);
+    for (int i = 0; i < size; i++) {
+      VERIFY_IS_EQUAL(result(i), accum);
+      accum *= tensor(i);
+    }
+}
+
 template <int DataLayout, typename Type=float>
 static void test_4d_scan()
 {