Chnage arma::square to square in methods

Signed-off-by: Omar Shrit <omar@avontech.fr>

src/mlpack/methods/ann/dists/normal_distribution_impl.hpp

@@ -45,8 +45,8 @@ DataType NormalDistribution<DataType>::LogProbability(
     const DataType& observation) const
 {
   const DataType v1 = arma::log(sigma) + std::log(std::sqrt(2 * M_PI));
-  const DataType v2 = arma::square(observation - mean) /
-                      (2 * arma::square(sigma));
+  const DataType v2 = square(observation - mean) /
+                      (2 * square(sigma));
   return  (-v1 - v2);
 }
 
@@ -56,9 +56,9 @@ void NormalDistribution<DataType>::ProbBackward(
     DataType& dmu,
     DataType& dsigma) const
 {
-  dmu = (observation - mean) / (arma::square(sigma)) % Probability(observation);
+  dmu = (observation - mean) / (square(sigma)) % Probability(observation);
   dsigma = (- 1.0 / sigma +
-            (arma::square(observation - mean) / arma::pow(sigma, 3)))
+            (square(observation - mean) / arma::pow(sigma, 3)))
             % Probability(observation);
 }
 

src/mlpack/methods/ann/layer/not_adapted/virtual_batch_norm_impl.hpp

@@ -38,7 +38,7 @@ VirtualBatchNormType<InputType, OutputType>::VirtualBatchNormType(
     loading(false)
 {
   referenceBatchMean = arma::mean(referenceBatch, 1);
-  referenceBatchMeanSquared = arma::mean(arma::square(referenceBatch), 1);
+  referenceBatchMeanSquared = arma::mean(square(referenceBatch), 1);
   newCoefficient = 1.0 / (referenceBatch.n_cols + 1);
   oldCoefficient = 1 - newCoefficient;
 }
@@ -68,12 +68,12 @@ void VirtualBatchNormType<InputType, OutputType>::Forward(
 
   inputParameter = input;
   InputType inputMean = arma::mean(input, 1);
-  InputType inputMeanSquared = arma::mean(arma::square(input), 1);
+  InputType inputMeanSquared = arma::mean(square(input), 1);
 
   mean = oldCoefficient * referenceBatchMean + newCoefficient * inputMean;
   OutputType meanSquared = oldCoefficient * referenceBatchMeanSquared +
       newCoefficient * inputMeanSquared;
-  variance = meanSquared - arma::square(mean);
+  variance = meanSquared - square(mean);
   // Normalize the input.
   output = input.each_col() - mean;
   inputSubMean = output;

src/mlpack/methods/ann/loss_functions/mean_squared_error_impl.hpp

@@ -30,7 +30,7 @@ typename MatType::elem_type MeanSquaredErrorType<MatType>::Forward(
     const MatType& target)
 {
   typename MatType::elem_type lossSum =
-      arma::accu(arma::square(prediction - target));
+      arma::accu(square(prediction - target));
 
   if (reduction)
     return lossSum;

src/mlpack/methods/ann/loss_functions/mean_squared_logarithmic_error_impl.hpp

@@ -31,7 +31,7 @@ typename MatType::elem_type MeanSquaredLogarithmicErrorType<MatType>::Forward(
     const MatType& target)
 {
   typename MatType::elem_type lossSum =
-      arma::accu(arma::square(arma::log(1.0 + target) -
+      arma::accu(square(arma::log(1.0 + target) -
       arma::log(1.0 + prediction)));
 
   if (reduction)

src/mlpack/methods/ann/not_adapted/rbm/spike_slab_rbm_impl.hpp

@@ -76,7 +76,7 @@ RBM<InitializationRuleType, DataType, PolicyType>::FreeEnergy(
 
   for (size_t i = 0; i < hiddenSize; ++i)
   {
-    ElemType sum = arma::accu(arma::square(input.t() * weight.slice(i))) /
+    ElemType sum = arma::accu(square(input.t() * weight.slice(i))) /
         (2.0 * slabPenalty);
     freeEnergy -= SoftplusFunction::Fn(spikeBias(i) - sum);
   }

src/mlpack/methods/decision_tree/mse_gain.hpp

@@ -70,7 +70,7 @@ class MSEGain
       Sum(values, begin, end, mean);
       mean /= (double) (end - begin);
 
-      mse = arma::accu(arma::square(values.subvec(begin, end - 1) - mean));
+      mse = arma::accu(square(values.subvec(begin, end - 1) - mean));
       mse /= (double) (end - begin);
     }
 
@@ -164,7 +164,7 @@ class MSEGain
 
     if (UseWeights)
     {
-      totalSumSquares = arma::accu(weights % arma::square(responses));
+      totalSumSquares = arma::accu(weights % square(responses));
       for (size_t i = 0; i < minimum - 1; ++i)
       {
         const WType w = weights[i];
@@ -192,7 +192,7 @@ class MSEGain
     }
     else
     {
-      totalSumSquares = arma::accu(arma::square(responses));
+      totalSumSquares = arma::accu(square(responses));
       for (size_t i = 0; i < minimum - 1; ++i)
       {
         const RType x = responses[i];

src/mlpack/methods/local_coordinate_coding/lcc_impl.hpp

@@ -329,7 +329,7 @@ inline double LocalCoordinateCoding::Objective(
     const size_t pointInd = (size_t) (adjacencies(l) / atoms);
 
     weightedL1NormZ += fabs(codes(atomInd, pointInd)) * arma::as_scalar(
-        arma::sum(arma::square(dictionary.col(atomInd) - data.col(pointInd))));
+        arma::sum(square(dictionary.col(atomInd) - data.col(pointInd))));
   }
 
   double froNormResidual = norm(data - dictionary * codes, "fro");

src/mlpack/methods/sparse_coding/sparse_coding_impl.hpp

@@ -182,7 +182,7 @@ inline double SparseCoding::OptimizeDictionary(const arma::mat& data,
 
     arma::mat matAInvZXT = solve(A, codesXT);
 
-    arma::vec gradient = -arma::sum(arma::square(matAInvZXT), 1);
+    arma::vec gradient = -arma::sum(square(matAInvZXT), 1);
     gradient += 1;
 
     arma::mat hessian = -(-2 * (matAInvZXT * trans(matAInvZXT)) % inv(A));