Make .clang-format competible with version 6

Clang-format 6 is widely used. However, `AllowAllConstructorInitializersOnNextLine` added from #203 is supported from clang-format 9. This Pr reverts use of `AllowAllConstructorInitializersOnNextLine` while having similar linting style. **Self evaluation:** 1. Build test: [X]Passed [ ]Failed [ ]Skipped 2. Run test: [X]Passed [ ]Failed [ ]Skipped Signed-off-by: Jihoon Lee <[email protected]>
nnstreamer · Jun 23, 2020 · a166b11 · a166b11
1 parent 260b691
commit a166b11
Show file tree

Hide file tree

Showing 18 changed files with 117 additions and 113 deletions.
diff --git a/.clang-format b/.clang-format
@@ -34,12 +34,12 @@ BraceWrapping:
 BreakBeforeBraces: Attach
 BreakBeforeTernaryOperators: true
 BreakConstructorInitializersBeforeComma: false
+BreakConstructorInitializers: AfterColon
 BreakAfterJavaFieldAnnotations: false
 BreakStringLiterals: true
 ColumnLimit:     80
 CommentPragmas:  '^ IWYU pragma:'
 ConstructorInitializerAllOnOneLineOrOnePerLine: true
-AllowAllConstructorInitializersOnNextLine: true
 ConstructorInitializerIndentWidth: 2
 ContinuationIndentWidth: 2
 Cpp11BracedListStyle: true

diff --git a/nntrainer/include/activation_layer.h b/nntrainer/include/activation_layer.h
@@ -25,7 +25,6 @@
 #include <vector>
 
 namespace nntrainer {
-
 
 /**
  * @class   Activation Layer
@@ -37,7 +36,7 @@ class ActivationLayer : public Layer {
   /**
    * @brief     Constructor of Activation Layer
    */
-  ActivationLayer() : Layer(){ this->type = LAYER_ACTIVATION; };
+  ActivationLayer() : Layer() { this->type = LAYER_ACTIVATION; };
 
   /**
    * @brief     Destructor of Activation Layer
@@ -57,13 +56,13 @@ class ActivationLayer : public Layer {
    * @brief     Read Activation layer params. This is essentially noops for now.
    * @param[in] file input stream file
    */
-  void read(std::ifstream &file) { /* noop */ };
+  void read(std::ifstream &file){/* noop */};
 
   /**
    * @brief     Save Activation layer params. This is essentially noops for now.
    * @param[in] file output stream file
    */
-  void save(std::ofstream &file) {  /* noop */ };
+  void save(std::ofstream &file){/* noop */};
 
   /**
    * @brief     forward propagation with input
@@ -109,9 +108,9 @@ class ActivationLayer : public Layer {
    *            activation_prime_function to be used
    * @retval #ML_ERROR_NONE when successful
    */
-  int
-  setActivation(std::function<float(float const)> const &activation_fn,
-                std::function<float(float const)> const &activation_prime_fn);
+  int setActivation(
+    std::function<float(float const)> const &activation_fn,
+    std::function<float(float const)> const &activation_prime_fn);
 
   /**
    * @brief setActivation by preset actiType

diff --git a/nntrainer/include/databuffer.h b/nntrainer/include/databuffer.h
@@ -116,13 +116,13 @@ class DataBuffer {
    * @brief     Create Buffer
    * @retval    DataBuffer
    */
-  DataBuffer()
-    : train_running(),
-      val_running(),
-      test_running(),
-      train_thread(),
-      val_thread(),
-      test_thread() {
+  DataBuffer() :
+    train_running(),
+    val_running(),
+    test_running(),
+    train_thread(),
+    val_thread(),
+    test_thread() {
     SET_VALIDATION(false);
     class_num = 0;
     cur_train_bufsize = 0;

diff --git a/nntrainer/include/layer.h b/nntrainer/include/layer.h
@@ -36,8 +36,10 @@ namespace nntrainer {
  * @brief     Enumeration of cost(loss) function type
  *            0. MSR ( Mean Squared Roots )
  *            1. ENTROPY ( Cross Entropy )
- *            2. ENTROPY_SIGMOID (Cross Entropy amalgamated with sigmoid for stability)
- *            2. ENTROPY_SOFTMAX (Cross Entropy amalgamated with softmax for stability)
+ *            2. ENTROPY_SIGMOID (Cross Entropy amalgamated with sigmoid for
+ * stability)
+ *            2. ENTROPY_SOFTMAX (Cross Entropy amalgamated with softmax for
+ * stability)
  *            2. Unknown
  */
 typedef enum {
@@ -115,18 +117,18 @@ typedef enum {
  */
 class Layer {
 public:
-  Layer()
-    : last_layer(false),
-      init_zero(false),
-      type(LAYER_UNKNOWN),
-      activation(NULL),
-      activation_prime(NULL),
-      loss(0.0),
-      cost(COST_UNKNOWN),
-      activation_type(ACT_UNKNOWN),
-      bn_follow(false),
-      weight_decay(),
-      weight_ini_type(WEIGHT_UNKNOWN) {}
+  Layer() :
+    last_layer(false),
+    init_zero(false),
+    type(LAYER_UNKNOWN),
+    activation(NULL),
+    activation_prime(NULL),
+    loss(0.0),
+    cost(COST_UNKNOWN),
+    activation_type(ACT_UNKNOWN),
+    bn_follow(false),
+    weight_decay(),
+    weight_ini_type(WEIGHT_UNKNOWN) {}
 
   /**
    * @brief     Destructor of Layer Class

diff --git a/nntrainer/include/lazy_tensor.h b/nntrainer/include/lazy_tensor.h
@@ -20,10 +20,11 @@
 
 #define FWD(...) std::forward<decltype(__VA_ARGS__)>(__VA_ARGS__)
 
-#define _LIFT(X)                                                  \
-  [](nntrainer::Tensor &t, auto &&... args) noexcept(             \
-    noexcept(t.X(FWD(args)...))) -> decltype(t.X(FWD(args)...)) { \
-    return t.X(FWD(args)...);                                     \
+#define _LIFT(X)                                            \
+  [](nntrainer::Tensor & t,                                 \
+     auto &&... args) noexcept(noexcept(t.X(FWD(args)...))) \
+    ->decltype(t.X(FWD(args)...)) {                         \
+    return t.X(FWD(args)...);                               \
   }
 
 namespace nntrainer {

diff --git a/nntrainer/include/loss_layer.h b/nntrainer/include/loss_layer.h
@@ -104,7 +104,7 @@ class LossLayer : public Layer {
    * @brief     update loss
    * @param[in] l Tensor data to calculate
    */
-  void updateLoss(const Tensor& l);
+  void updateLoss(const Tensor &l);
 };
 } // namespace nntrainer
 

diff --git a/nntrainer/include/neuralnet.h b/nntrainer/include/neuralnet.h
@@ -336,7 +336,8 @@ class NeuralNetwork {
   std::shared_ptr<DataBuffer> data_buffer;
 
   /**
-   * @brief    Continue train from the previous state of optimizer and iterations
+   * @brief    Continue train from the previous state of optimizer and
+   * iterations
    */
   bool continue_train;
 

diff --git a/nntrainer/include/optimizer.h b/nntrainer/include/optimizer.h
@@ -64,10 +64,16 @@ typedef struct _OptParam {
   double epsilon;
   float decay_rate;
   float decay_steps;
-  bool continue_train;  /** Continue training with previous tensors for adam */
-
-  _OptParam() : learning_rate(0.0), beta1(0.0), beta2(0.0), epsilon(0.0),
-    decay_rate(0.0), decay_steps(0.0), continue_train(false) {}
+  bool continue_train; /** Continue training with previous tensors for adam */
+
+  _OptParam() :
+    learning_rate(0.0),
+    beta1(0.0),
+    beta2(0.0),
+    epsilon(0.0),
+    decay_rate(0.0),
+    decay_steps(0.0),
+    continue_train(false) {}
 } OptParam;
 
 class Optimizer {

diff --git a/nntrainer/include/pooling2d_layer.h b/nntrainer/include/pooling2d_layer.h
@@ -119,7 +119,6 @@ class Pooling2DLayer : public Layer {
    */
   int setProperty(std::vector<std::string> values);
 
-
   /* TO DO : support keras type of padding */
   enum class PaddingType {
     full = 0,
@@ -143,7 +142,6 @@ class Pooling2DLayer : public Layer {
    * @retval Tensor outoput tensor
    */
   Tensor pooling2d(unsigned int batch, Tensor in, int &status);
-
 };
 
 } // namespace nntrainer

diff --git a/nntrainer/src/fc_layer.cpp b/nntrainer/src/fc_layer.cpp
@@ -192,12 +192,11 @@ Tensor FullyConnectedLayer::backwarding(Tensor derivative, int iteration) {
 
     case COST_ENTROPY:
       throw std::runtime_error(
-          "Error: Cross Entropy not supported without softmax or sigmoid.");
+        "Error: Cross Entropy not supported without softmax or sigmoid.");
     case COST_UNKNOWN:
       /** Intended */
     default:
-      throw std::runtime_error(
-          "Error: unknown cost.");
+      throw std::runtime_error("Error: unknown cost.");
     }
   }
 

diff --git a/nntrainer/src/layer.cpp b/nntrainer/src/layer.cpp
@@ -107,16 +107,15 @@ Tensor Layer::initializeWeight(TensorDim w_dim, WeightIniType init_type,
     w.setRandNormal(0, sqrt(2.0 / (dim.height())));
     break;
   case WEIGHT_LECUN_UNIFORM:
-    w.setRandUniform(-1.0 * sqrt(1.0 / dim.height()),
-        sqrt(1.0 / dim.height()));
+    w.setRandUniform(-1.0 * sqrt(1.0 / dim.height()), sqrt(1.0 / dim.height()));
     break;
   case WEIGHT_XAVIER_UNIFORM:
     w.setRandUniform(-1.0 * sqrt(6.0 / (dim.height() + dim.width())),
-        sqrt(6.0 / (dim.height() + dim.width())));
+                     sqrt(6.0 / (dim.height() + dim.width())));
     break;
   case WEIGHT_HE_UNIFORM:
     w.setRandUniform(-1.0 * sqrt(6.0 / (dim.height())),
-        sqrt(6.0 / (dim.height())));
+                     sqrt(6.0 / (dim.height())));
     break;
   default:
     break;

diff --git a/nntrainer/src/loss_layer.cpp b/nntrainer/src/loss_layer.cpp
@@ -21,9 +21,9 @@
  *
  */
 
-#include <loss_layer.h>
 #include <layer.h>
 #include <lazy_tensor.h>
+#include <loss_layer.h>
 #include <nntrainer_error.h>
 #include <nntrainer_log.h>
 #include <parse_util.h>
@@ -67,12 +67,15 @@ Tensor LossLayer::forwarding(Tensor output, Tensor label, int &status) {
     // @note: the output should be logit before applying sigmoid
     // log(1 + exp(-abs(y))) + max(y, 0)
     Tensor mid_term = y.apply(static_cast<float (*)(float)>(&std::fabs))
-      .multiply(-1.0).apply(static_cast<float (*)(float)>(&std::exp))
-      .add(1.0).apply(logFloat);
+                        .multiply(-1.0)
+                        .apply(static_cast<float (*)(float)>(&std::exp))
+                        .add(1.0)
+                        .apply(logFloat);
     mid_term = mid_term.add(mid_term.apply(relu));
 
     // loss = y * y2 - (log(1 + exp(-abs(y))) + max(y, 0))
-    l = y2.chain().multiply_i(y)
+    l = y2.chain()
+          .multiply_i(y)
           .add_i(mid_term)
           .multiply_i(-1.0 / y2.getWidth())
           .run()
@@ -106,7 +109,7 @@ Tensor LossLayer::forwarding(Tensor output, Tensor label, int &status) {
   return y;
 }
 
-void LossLayer::updateLoss(const Tensor& l) {
+void LossLayer::updateLoss(const Tensor &l) {
   float loss_sum = 0.0;
   const float *data = l.getData();
 
@@ -117,8 +120,7 @@ void LossLayer::updateLoss(const Tensor& l) {
 }
 
 void LossLayer::copy(std::shared_ptr<Layer> l) {
-  std::shared_ptr<LossLayer> from =
-    std::static_pointer_cast<LossLayer>(l);
+  std::shared_ptr<LossLayer> from = std::static_pointer_cast<LossLayer>(l);
   this->last_layer = from->last_layer;
   this->input.copy(from->input);
   this->cost = from->cost;
@@ -144,7 +146,7 @@ Tensor LossLayer::backwarding(Tensor derivative, int iteration) {
     break;
   case COST_ENTROPY:
     throw std::runtime_error(
-        "Error: Cross Entropy not supported without softmax or sigmoid.");
+      "Error: Cross Entropy not supported without softmax or sigmoid.");
   case COST_UNKNOWN:
     /** intended */
   default:

diff --git a/nntrainer/src/neuralnet.cpp b/nntrainer/src/neuralnet.cpp
@@ -93,7 +93,8 @@ std::vector<std::string> parseLayerName(std::string ll) {
 
 NeuralNetwork::NeuralNetwork() : NeuralNetwork("") {}
 
-NeuralNetwork::NeuralNetwork(std::string config) : batch_size(0),
+NeuralNetwork::NeuralNetwork(std::string config) :
+  batch_size(0),
   learning_rate(0.0),
   decay_rate(0.0),
   decay_steps(0.0),
@@ -294,26 +295,23 @@ int NeuralNetwork::init() {
                     ini, (layers_name[i] + ":kernel_size").c_str(), unknown),
                   (int *)size);
       NN_INI_RETURN_STATUS();
-      status = conv2d_layer->setSize(
-        size, Layer::PropertyType::kernel_size);
+      status = conv2d_layer->setSize(size, Layer::PropertyType::kernel_size);
       NN_INI_RETURN_STATUS();
 
       status = getValues(
         CONV2D_DIM,
         iniparser_getstring(ini, (layers_name[i] + ":stride").c_str(), unknown),
         (int *)size);
       NN_INI_RETURN_STATUS();
-      status = conv2d_layer->setSize(
-        size, Layer::PropertyType::stride);
+      status = conv2d_layer->setSize(size, Layer::PropertyType::stride);
       NN_INI_RETURN_STATUS();
 
       status = getValues(CONV2D_DIM,
                          iniparser_getstring(
                            ini, (layers_name[i] + ":padding").c_str(), unknown),
                          (int *)size);
       NN_INI_RETURN_STATUS();
-      status = conv2d_layer->setSize(
-        size, Layer::PropertyType::padding);
+      status = conv2d_layer->setSize(size, Layer::PropertyType::padding);
       NN_INI_RETURN_STATUS();
 
       status = conv2d_layer->setFilter(
@@ -586,7 +584,7 @@ int NeuralNetwork::init(std::shared_ptr<Optimizer> optimizer,
       status = fc_layer->setOptimizer(opt);
       NN_RETURN_STATUS();
 
-    }  break;
+    } break;
     case LAYER_BN:
       layers[i]->setInputDimension(previous_dim);
       status = layers[i]->initialize(last);
@@ -643,12 +641,12 @@ Tensor NeuralNetwork::forwarding(Tensor input, Tensor output, int &status) {
   Tensor X = input;
   Tensor Y2 = output;
 
-  X = forwarding (input, status);
+  X = forwarding(input, status);
   if (status != ML_ERROR_NONE)
     return X;
 
   X = std::static_pointer_cast<LossLayer>(layers[layers.size() - 1])
-      ->forwarding(X, Y2, status);
+        ->forwarding(X, Y2, status);
   return X;
 }
 
@@ -674,7 +672,7 @@ int NeuralNetwork::backwarding(Tensor input, Tensor expected_output,
 
 float NeuralNetwork::getLoss() {
   loss = 0.0;
-  for (unsigned int i=0; i < layers.size(); i++) {
+  for (unsigned int i = 0; i < layers.size(); i++) {
     loss += layers[i]->getLoss();
   }
 
@@ -853,7 +851,7 @@ int NeuralNetwork::train_run() {
           backwarding(nntrainer::Tensor(in), nntrainer::Tensor(label), iter++);
         if (status != ML_ERROR_NONE) {
           data_buffer->clear(nntrainer::BUF_TRAIN);
-          ml_loge ("Error: training error in #%d/%d.", i+1, epoch);
+          ml_loge("Error: training error in #%d/%d.", i + 1, epoch);
           return status;
         }
         std::cout << "#" << i + 1 << "/" << epoch;
@@ -889,7 +887,7 @@ int NeuralNetwork::train_run() {
             nntrainer::Tensor Y2 = nntrainer::Tensor({label[i]});
             nntrainer::Tensor Y = forwarding(X, Y2, status);
             if (status != ML_ERROR_NONE) {
-              ml_loge ("Error: forwarding the network resulted in error.");
+              ml_loge("Error: forwarding the network resulted in error.");
               return status;
             }