25 references to NormSquared
Microsoft.ML.Core.Tests (1)
UnitTests\TestVBuffer.cs (1)
105Assert.True(CompareNumbersWithTolerance(l2Squared, VectorUtils.NormSquared(in a), digitsOfPrecision: tol));
Microsoft.ML.Data (1)
Deprecated\Vector\VBufferMathUtils.cs (1)
41return MathUtils.Sqrt(NormSquared(in a));
Microsoft.ML.KMeansClustering (11)
KMeansModelParameters.cs (2)
169float instanceL2 = VectorUtils.NormSquared(in src); 270_centroidL2s[i] = VectorUtils.NormSquared(_centroids[i]);
KMeansPlusPlusTrainer.cs (9)
394l2 = VectorUtils.NormSquared(cursor.Features); 441centroidL2s[i] = cachedCandidateL2 ?? VectorUtils.NormSquared(candidate); 699float pointNorm = VectorUtils.NormSquared(in point); 860clustersL2s[clusterCount] = VectorUtils.NormSquared(clusters[clusterCount]); 896clustersL2s[clusterCount] = VectorUtils.NormSquared(clusters[clusterCount]); 1225centroidL2s[i] = VectorUtils.NormSquared(Centroids[i]); 1305float instanceNormSquared = VectorUtils.NormSquared(in features); 1363centroidL2s[i] = VectorUtils.NormSquared(in centroids[i]); 1827float l2 = VectorUtils.NormSquared(in features);
Microsoft.ML.PCA (3)
PcaTrainer.cs (3)
470_norm2Mean = VectorUtils.NormSquared(mean); 497_norm2Mean = VectorUtils.NormSquared(_mean); 637float norm2X = VectorUtils.NormSquared(in src) -
Microsoft.ML.StandardTrainers (9)
Optimizer\OptimizationMonitor.cs (1)
323float gradientNormSquared = VectorUtils.NormSquared(gradient);
Optimizer\SgdOptimizer.cs (2)
341float newByNew = VectorUtils.NormSquared(_newGrad); 343float oldByOld = VectorUtils.NormSquared(_grad);
Standard\SdcaBinary.cs (4)
583var normSquared = VectorUtils.NormSquared(features); 824var featuresNormSquared = VectorUtils.NormSquared(features); 997var l2Regularizer = l2Const * (VectorUtils.NormSquared(weights[0]) + biasReg[0] * biasReg[0]) * 0.5; 2094var newLoss = lossSum.Sum / count + l2Weight * VectorUtils.NormSquared(weights) * 0.5;
Standard\SdcaMulticlass.cs (2)
218normSquared = VectorUtils.NormSquared(in features); 422weightsL2NormSquared += VectorUtils.NormSquared(weights[iClass]) + biasReg[iClass] * biasReg[iClass];