Schema

Data\RoleMappedSchema.cs (1)

Data\RoleMappedSchema.cs (2)

405/// Note that the schema of <see cref="RoleMappedSchema.Schema"/> of <see cref="Schema"/> is 417/// The role mapped schema. Note that <see cref="Schema"/>'s <see cref="RoleMappedSchema.Schema"/> is

Commands\CrossValidationCommand.cs (8)

259return new RoleMappedData(pipe, srcData.Schema.GetColumnRoleNames()); 531var mapper = bindable.Bind(host, testData.Schema); 533IDataScorerTransform scorePipe = scorerComp.CreateComponent(host, testData.Data, mapper, trainData.Schema); 544trainData.Schema.GetColumnRoleNames()); 555var dataEval = new RoleMappedData(scorePipe, testData.Schema.GetColumnRoleNames(), opt: true); 562perInstance = new RoleMappedData(perInst, dataEval.Schema.GetColumnRoleNames(), opt: true); 564return new FoldResult(dict, dataEval.Schema.Schema, perInstance, trainData.Schema);

Commands\ScoreCommand.cs (1)

243var sc = GetScorerComponentAndMapper(predictor, null, data.Schema, env, null, out var mapper);

Commands\TrainCommand.cs (5)

178validData = new RoleMappedData(validPipe, data.Schema.GetColumnRoleNames()); 196testDataUsedInTrainer = new RoleMappedData(testPipeUsedInTrainer, data.Schema.GetColumnRoleNames()); 377ModelFileUtils.SaveRoleMappings(env, ch, data.Schema, rep); 503var prefetch = data.Schema.GetColumnRoles().Select(kc => kc.Value.Index).ToArray(); 506data = new RoleMappedData(cacheView, data.Schema.GetColumnRoleNames());

Commands\TrainTestCommand.cs (3)

164validData = new RoleMappedData(validPipe, data.Schema.GetColumnRoleNames()); 182testDataUsedInTrainer = new RoleMappedData(testPipeUsedInTrainer, data.Schema.GetColumnRoleNames()); 205IDataScorerTransform scorePipe = ScoreUtils.GetScorer(ImplOptions.Scorer, predictor, testPipe, features, group, customCols, Host, data.Schema);

EntryPoints\InputBase.cs (1)

116cachedRoleMappedData = new RoleMappedData(outputData, roleMappedData.Schema.GetColumnRoleNames());

EntryPoints\PredictorModelImpl.cs (1)

37_roleMappings = trainingData.Schema.GetColumnRoleNames().ToArray();

EntryPoints\SummarizePredictor.cs (1)

44output.Summary = GetSummaryAndStats(host, predictor, rmd.Schema, out output.Stats);

Evaluators\EvaluatorBase.cs (6)

33CheckColumnTypes(data.Schema); 34Func<int, bool> activeCols = GetActiveCols(data.Schema); 35var agg = GetAggregator(data.Schema); 36AggregatorDictionaryBase[] dictionaries = GetAggregatorDictionaries(data.Schema); 38var dict = ProcessData(data.Data, data.Schema, activeCols, agg, dictionaries); 459var mapper = CreatePerInstanceRowMapper(data.Schema);

Evaluators\EvaluatorUtils.cs (1)

774var combined = AppendPerInstanceDataViews(env, perInstance[0].Schema.Label?.Name, foldDataViews, out variableSizeVectorColumnNames);

Evaluators\MamlEvaluator.cs (9)

19/// evaluation should be searched for by name in the <see cref="RoleMappedData.Schema"/>. 109data = new RoleMappedData(data.Data, GetInputColumnRoles(data.Schema, needStrat: true)); 222var schema = scoredData.Schema; 238if (perInst.Schema.Schema.TryGetColumnIndex(MetricKinds.ColumnNames.FoldIndex, out foldCol)) 242if (perInst.Schema.Name?.Name is string nameName) 257if (perInst.Schema.Weight?.Name is string weightName) 261foreach (var col in GetPerInstanceColumnsToSave(perInst.Schema)) 266return GetPerInstanceMetricsCore(idv, perInst.Schema); 284var data = new RoleMappedData(perInstance.Data, GetInputColumnRoles(perInstance.Schema, needName: true));

Evaluators\RankingEvaluator.cs (5)

137Host.CheckParam(data.Schema.Label.HasValue, nameof(data), "Schema must contain a label column"); 138var scoreInfo = data.Schema.GetUniqueColumn(AnnotationUtils.Const.ScoreValueKind.Score); 139Host.CheckParam(data.Schema.Group.HasValue, nameof(data), "Schema must contain a group column"); 142data.Schema.Label.Value.Name, scoreInfo.Name, data.Schema.Group.Value.Name, _truncationLevel, _labelGains);

Prediction\Calibrator.cs (5)

916if (!NeedCalibration(env, ch, calibrator, trainer, predictor, data.Schema)) 1020ch.CheckParam(data.Schema.Label.HasValue, nameof(data), "data must have a Label column"); 1024return TrainCalibrator(env, ch, caliTrainer, scored, data.Schema.Label.Value.Name, DefaultColumnNames.Score, data.Schema.Weight?.Name, maxRows); 2222if (data.Schema.Label == null && scp != null)

Training\TrainerUtils.cs (33)

53if (!data.Schema.Feature.HasValue) 55var col = data.Schema.Feature.Value; 70Contracts.Assert(data.Schema.Feature.HasValue); 71var col = data.Schema.Feature.Value; 86if (!data.Schema.Label.HasValue) 88var col = data.Schema.Label.Value; 121if (!data.Schema.Label.HasValue) 123var col = data.Schema.Label.Value; 124Contracts.Assert(!data.Schema.Schema[col.Index].IsHidden); 141if (!data.Schema.Label.HasValue) 143var col = data.Schema.Label.Value; 189if (!data.Schema.Label.HasValue) 191var col = data.Schema.Label.Value; 203if (!data.Schema.Weight.HasValue) 205var col = data.Schema.Weight.Value; 215if (!data.Schema.Group.HasValue) 217var col = data.Schema.Group.Value; 233if ((opt & CursOpt.Label) != 0 && data.Schema.Label.HasValue) 234columns.Add(data.Schema.Label.Value); 235if ((opt & CursOpt.Features) != 0 && data.Schema.Feature.HasValue) 236columns.Add(data.Schema.Feature.Value); 237if ((opt & CursOpt.Weight) != 0 && data.Schema.Weight.HasValue) 238columns.Add(data.Schema.Weight.Value); 239if ((opt & CursOpt.Group) != 0 && data.Schema.Group.HasValue) 240columns.Add(data.Schema.Group.Value); 278return GetFeatureFloatVectorGetter(row, data.Schema); 302return GetLabelFloatGetter(row, data.Schema); 323return GetOptWeightFloatGetter(row, data.Schema); 344return GetOptGroupGetter(row, data.Schema); 427var tschema = context.TrainingSet.Schema; 834if ((opt & CursOpt.Features) != 0 && data.Schema.Feature != null) 901if ((opt & CursOpt.Label) != 0 && data.Schema.Label != null) 973if ((opt & CursOpt.Label) != 0 && data.Schema.Label != null)

Transforms\TrainAndScoreTransformer.cs (1)

245return ScoreUtils.GetScorer(args.Scorer, predictor, input, feat, group, customCols, env, data.Schema, mapperFactory);

Utilities\ComponentCreation.cs (1)

172var mapper = bindable.Bind(env, data.Schema);

EnsembleUtils.cs (3)

22Contracts.Assert(data.Schema.Feature.HasValue); 24var featCol = data.Schema.Feature.Value; 39var res = new RoleMappedData(view, data.Schema.GetColumnRoleNames());

EntryPoints\PipelineEnsemble.cs (1)

54summaries[i] = SummarizePredictor.GetSummaryAndStats(host, pred, rmd.Schema, out stats[i]);

OutputCombiners\BaseStacking.cs (3)

155switch (data.Schema.Label.Value.Type.GetRawKind()) 162ch.Check(data.Schema.Label.Value.Type is KeyDataViewType); 208bldr.AddColumn(DefaultColumnNames.Label, data.Schema.Label.Value.Type as PrimitiveDataViewType, labels);

PipelineEnsemble.cs (10)

72Mappers[i] = bindable.Bind(Parent.Host, rmd.Schema) as ISchemaBoundRowMapper; 584summaryModel.SaveSummary(writer, rmd.Schema); 597if (!rmd.Schema.Label.HasValue) 599var labelCol = rmd.Schema.Label.Value; 607var schema = rmd.Schema.Schema; 660var labelInfo = rmd.Schema.Label.HasValue; 661if (!rmd.Schema.Label.HasValue) 663var labelCol = rmd.Schema.Label.Value; 727summaryModel.SaveSummary(sw, rmd.Schema); 732var listCur = summaryKvps.GetSummaryInKeyValuePairs(rmd.Schema);

Selector\FeatureSelector\RandomFeatureSelector.cs (1)

52var type = data.Schema.Feature.Value.Type;

Selector\SubModelSelector\BaseSubModelSelector.cs (2)

83IDataScorerTransform scorePipe = ScoreUtils.GetScorer(model.Predictor, testData, Host, testData.Schema); 85GetColumnRoles(testData.Schema, scorePipe.Schema));

Selector\SubsetSelector\BaseSubsetSelector.cs (2)

80dataTest = new RoleMappedData(viewTest, Data.Schema.GetColumnRoleNames()); 81dataTrain = new RoleMappedData(viewTrain, Data.Schema.GetColumnRoleNames());

Selector\SubsetSelector\BootstrapSelector.cs (1)

52var dataTrain = new RoleMappedData(viewTrain, Data.Schema.GetColumnRoleNames());

Selector\SubsetSelector\RandomPartitionSelector.cs (1)

50var dataTrain = new RoleMappedData(viewTrain, Data.Schema.GetColumnRoleNames());

PermutationFeatureImportance.cs (11)

ScoreModel.cs (2)

85var mapper = bindable.Bind(host, data.Schema); 135var mapper = bindable.Bind(host, data.Schema);

FastTree.cs (19)

190AnnotationUtils.TryGetCategoricalFeatureIndices(trainData.Schema.Schema, trainData.Schema.Feature.Value.Index, out CategoricalFeatures); 205Host.Assert(data.Schema.Feature.HasValue); 211return itdv?.GetSlotType(data.Schema.Feature.Value.Index) != null; 1329Host.Assert(examples.Schema.Feature.HasValue); 1345var labelName = examples.Schema.Label?.Name; 1354if (examples.Schema.Group?.Name is string groupName) 1359examples = new RoleMappedData(data, examples.Schema.GetColumnRoleNames()); 1364int featIdx = AddColumnIfNeeded(examples.Schema.Feature, toTranspose); 1365int labelIdx = AddColumnIfNeeded(examples.Schema.Label, toTranspose); 1366int groupIdx = AddColumnIfNeeded(examples.Schema.Group, toTranspose); 1367int weightIdx = AddColumnIfNeeded(examples.Schema.Weight, toTranspose); 1377FeaturesToContentMap fmap = new FeaturesToContentMap(examples.Schema); 1618ch.Warning("This is not ranking problem, Group Id '{0}' column will be ignored", examples.Schema.Group.Value.Name); 1771_weights = data.Schema.Weight != null ? new List<double>() : null; 1824hasGroup = _data.Schema.Group != null; 1831if (_data.Schema.Group != null) 1832ch.Warning("This is not ranking problem, Group Id '{0}' column will be ignored", _data.Schema.Group.Value.Name); 1835if (_data.Schema.Weight.HasValue)

FastTreeClassification.cs (1)

197FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount();

FastTreeRanking.cs (1)

150FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount();

FastTreeRegression.cs (1)

120FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount();

FastTreeTweedie.cs (1)

130FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount();

GamModelParameters.cs (3)

633var schema = _data.Schema; 687var featureCol = _data.Schema.Schema[DefaultColumnNames.Features]; 688AnnotationUtils.TryGetCategoricalFeatureIndices(_data.Schema.Schema, featureCol.Index, out _catsMap);

GamTrainer.cs (3)

233InputLength = context.TrainingSet.Schema.Feature.Value.Type.GetValueCount(); 270Host.Assert(data.Schema.Feature.HasValue); 274return (data.Data as ITransposeDataView)?.GetSlotType(data.Schema.Feature.Value.Index) != null;

RandomForestClassification.cs (2)

224FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount(); 227if (!trainData.Schema.Weight.HasValue && MLContext.OneDalDispatchingEnabled)

RandomForestRegression.cs (2)

363FeatureCount = trainData.Schema.Feature.Value.Type.GetValueCount(); 366if (!trainData.Schema.Weight.HasValue && MLContext.OneDalDispatchingEnabled)

TreeEnsembleFeaturizer.cs (10)

646Contracts.Assert(data.Schema.Feature.HasValue); 655if (vm.InputType.GetVectorSize() != data.Schema.Feature.Value.Type.GetVectorSize()) 659vm.InputType.GetVectorSize(), data.Schema.Feature.Value.Type.GetVectorSize()); 663var bound = bindable.Bind(env, data.Schema); 664xf = new GenericScorer(env, scorerArgs, input, bound, data.Schema); 716ch.Assert(data.Schema.Feature.HasValue); 726if (data != null && vm.InputType.GetVectorSize() != data.Schema.Feature.Value.Type.GetVectorSize()) 730vm.InputType.GetVectorSize(), data.Schema.Feature.Value.Type.GetVectorSize()); 734var bound = bindable.Bind(env, data.Schema); 735return new GenericScorer(env, scorerArgs, data.Data, bound, data.Schema);

KMeansPlusPlusTrainer.cs (1)

230if (data.Schema.Weight.HasValue)

LightGbmBinaryTrainer.cs (2)

270var labelType = data.Schema.Label.Value.Type; 274$"Label column '{data.Schema.Label.Value.Name}' is of type '{labelType.RawType}', but must be unsigned int, boolean or float.");

LightGbmMulticlassTrainer.cs (3)

240var labelType = data.Schema.Label.Value.Type; 244$"Label column '{data.Schema.Label.Value.Name}' is of type '{labelType.RawType}', but must be of unsigned int, boolean or float."); 293if (data.Schema.Label.Value.Type is KeyDataViewType keyType)

LightGbmRankingTrainer.cs (4)

248var labelCol = data.Schema.Label.Value; 256if (!data.Schema.Group.HasValue) 257throw ch.ExceptValue(nameof(data.Schema.Group), "Group column is missing."); 258var groupCol = data.Schema.Group.Value;

LightGbmRegressionTrainer.cs (2)

232var labelType = data.Schema.Label.Value.Type; 236$"Label column '{data.Schema.Label.Value.Name}' is of type '{labelType.RawType}', but must be an unsigned int, boolean or float.");

LightGbmTrainerBase.cs (8)

526ch.CheckParam(data.Schema.Label.HasValue, nameof(data), "Need a label column"); 557if (data.Schema.Weight.HasValue) 645var featureCol = trainData.Schema.Feature.Value; 646AnnotationUtils.TryGetCategoricalFeatureIndices(trainData.Schema.Schema, featureCol.Index, out categoricalFeatures); 648var colType = trainData.Schema.Feature.Value.Type; 769ch.Check(factory.Data.Schema.Label != null, "The data should have label."); 771bool hasWeights = factory.Data.Schema.Weight != null; 775ch.Check(factory.Data.Schema.Group != null, "The data for ranking task should have group field.");

OlsLinearRegression.cs (5)

157ch.CheckParam(examples.Schema.Feature.HasValue, nameof(examples), "Need a feature column"); 158ch.CheckParam(examples.Schema.Label.HasValue, nameof(examples), "Need a labelColumn column"); 161var typeLab = examples.Schema.Label.Value.Type; 166var typeFeat = examples.Schema.Feature.Value.Type as VectorDataViewType; 173if (examples.Schema.Weight.HasValue)

SymSgdClassificationTrainer.cs (7)

205var roles = examples.Schema.GetColumnRoleNames(); 208ch.Assert(examplesToFeedTrain.Schema.Label.HasValue); 209ch.Assert(examplesToFeedTrain.Schema.Feature.HasValue); 210if (examples.Schema.Weight.HasValue) 211ch.Assert(examplesToFeedTrain.Schema.Weight.HasValue); 213ch.Check(examplesToFeedTrain.Schema.Feature.Value.Type is VectorDataViewType vecType && vecType.Size > 0, "Training set has no features, aborting training."); 703int numFeatures = data.Schema.Feature.Value.Type.GetVectorSize();

PcaTrainer.cs (1)

231if (data.Schema.Weight.HasValue)

MatrixFactorizationTrainer.cs (8)

441ch.CheckParam(data.Schema.Label.HasValue, nameof(data), "Input data did not have a unique label"); 443var labelCol = data.Schema.Label.Value; 450ch.CheckParam(validData.Schema.Label.HasValue, nameof(validData), "Input validation data did not have a unique label"); 452var validLabelCol = validData.Schema.Label.Value; 474using (var cursor = data.Data.GetRowCursor(matrixColumnIndexColInfo, matrixRowIndexColInfo, data.Schema.Label.Value)) 477var labGetter = RowCursorUtils.GetGetterAs<float>(NumberDataViewType.Single, cursor, data.Schema.Label.Value.Index); 493using (var validCursor = validData.Data.GetRowCursor(matrixColumnIndexColInfo, matrixRowIndexColInfo, data.Schema.Label.Value)) 495ValueGetter<float> validLabelGetter = RowCursorUtils.GetGetterAs<float>(NumberDataViewType.Single, validCursor, validData.Schema.Label.Value.Index);

RecommenderUtils.cs (3)

54if (!data.Schema.HasUnique(role)) 56int kindCount = Utils.Size(data.Schema.GetColumns(role)); 61col = data.Schema.GetColumns(role)[0];

FactorizationMachine\FactorizationMachineTrainer.cs (15)

362var featureColumns = data.Schema.GetColumns(RoleMappedSchema.ColumnRole.Feature); 373columns.Add(data.Schema.Label.Value); 374if (data.Schema.Weight != null) 375columns.Add(data.Schema.Weight.Value); 379var labelGetter = RowCursorUtils.GetLabelGetter(cursor, data.Schema.Label.Value.Index); 380var weightGetter = data.Schema.Weight.HasValue ? cursor.GetGetter<float>(data.Schema.Weight.Value) : null; 415var featureColumns = data.Schema.GetColumns(RoleMappedSchema.ColumnRole.Feature); 440var validFeatureColumns = data.Schema.GetColumns(RoleMappedSchema.ColumnRole.Feature); 478var columns = data.Schema.Schema.Where(x => fieldColumnIndexes.Contains(x.Index)).ToList(); 479columns.Add(data.Schema.Label.Value); 480if (data.Schema.Weight != null) 481columns.Add(data.Schema.Weight.Value); 491var labelGetter = RowCursorUtils.GetLabelGetter(cursor, data.Schema.Label.Value.Index); 492var weightGetter = data.Schema.Weight?.Index is int weightIdx ? RowCursorUtils.GetGetterAs<float>(NumberDataViewType.Single, cursor, weightIdx) : null;

LdSvm\LdSvmTrainer.cs (4)

591using (var cursor = _data.Data.GetRowCursor(_data.Data.Schema[_data.Schema.Feature.Value.Name])) 593var getter = cursor.GetGetter<VBuffer<float>>(_data.Data.Schema[_data.Schema.Feature.Value.Name]); 632var featureCol = _data.Data.Schema[_data.Schema.Feature.Value.Name]; 633var labelCol = _data.Data.Schema[_data.Schema.Label.Value.Name];

Standard\LogisticRegression\LbfgsPredictorBase.cs (4)

481if (data.Schema.Weight.HasValue) 487var typeFeat = data.Schema.Feature.Value.Type as VectorDataViewType; 587if (data.Schema.Weight != null) 592if (data.Schema.Weight.HasValue)

Standard\LogisticRegression\LogisticRegression.cs (1)

268var featureCol = cursorFactory.Data.Schema.Feature.Value;

Standard\LogisticRegression\MulticlassLogisticRegression.cs (1)

181var labelCol = data.Schema.Label.Value;

Standard\MulticlassClassification\MetaMulticlassTrainer.cs (2)

97Host.Assert(data.Schema.Label.HasValue); 99var label = data.Schema.Label.Value;

Standard\MulticlassClassification\MulticlassNaiveBayesTrainer.cs (3)

136Host.Check(data.Schema.Label.HasValue, "Missing Label column"); 137var labelCol = data.Schema.Label.Value; 141Host.Check(data.Schema.Feature.HasValue, "Missing Feature column");

Standard\MulticlassClassification\OneVersusAllTrainer.cs (2)

167string trainerLabel = data.Schema.Label.Value.Name; 193var label = data.Schema.Label.Value;

Standard\MulticlassClassification\PairwiseCouplingTrainer.cs (2)

149string trainerLabel = data.Schema.Label.Value.Name; 166var label = data.Schema.Label.Value;

Standard\Online\OnlineLinear.cs (1)

324if (data.Schema.Weight.HasValue)

Standard\SdcaBinary.cs (11)

116var roles = examples.Schema.GetColumnRoleNames(); 119ch.Assert(examplesToFeedTrain.Schema.Label.HasValue); 120ch.Assert(examplesToFeedTrain.Schema.Feature.HasValue); 121if (examples.Schema.Weight.HasValue) 122ch.Assert(examplesToFeedTrain.Schema.Weight.HasValue); 124ch.Check(examplesToFeedTrain.Schema.Feature.Value.Type is VectorDataViewType vecType && vecType.Size > 0, "Training set has no features, aborting training."); 327int numFeatures = data.Schema.Feature.Value.Type.GetVectorSize(); 331if (data.Schema.Weight.HasValue) 2024Contracts.Assert(data.Schema.Feature.HasValue); 2026int numFeatures = data.Schema.Feature.Value.Type.GetVectorSize(); 2029if (data.Schema.Weight.HasValue)

Standard\Simple\SimpleTrainers.cs (7)

255_host.CheckParam(data.Schema.Label.HasValue, nameof(data), "Missing Label column"); 256var labelCol = data.Schema.Label.Value; 263if (data.Schema.Weight?.Type == NumberDataViewType.Single) 264colWeight = data.Schema.Weight.Value.Index; 266var cols = colWeight > -1 ? new DataViewSchema.Column[] { labelCol, data.Schema.Weight.Value } : new DataViewSchema.Column[] { labelCol }; 270var getLab = cursor.GetGetter<bool>(data.Schema.Label.Value); 271var getWeight = colWeight >= 0 ? cursor.GetGetter<float>(data.Schema.Weight.Value) : null;

Standard\StochasticTrainerBase.cs (6)

80var roles = examples.Schema.GetColumnRoleNames(); 83ch.Assert(examplesToFeedTrain.Schema.Label.HasValue); 84ch.Assert(examplesToFeedTrain.Schema.Feature.HasValue); 85if (examples.Schema.Weight.HasValue) 86ch.Assert(examplesToFeedTrain.Schema.Weight.HasValue); 88ch.Check(examplesToFeedTrain.Schema.Feature.Value.Type is VectorDataViewType vecType && vecType.Size > 0, "Training set has no features, aborting training.");

AdaptiveSingularSpectrumSequenceModeler.cs (6)

1210_host.CheckParam(data.Schema.Feature.HasValue, nameof(data), "Must have features column."); 1211var featureCol = data.Schema.Feature.Value; 1557if (data.Schema.Feature.Value.Type != NumberDataViewType.Single) 1558throw _host.ExceptUserArg(nameof(data.Schema.Feature.Value.Name), "The time series input column has " + 1559"type '{0}', but must be a float.", data.Schema.Feature.Value.Type); 1561var col = data.Schema.Feature.Value;

Dracula\CountTableTransformer.cs (2)

298_host.Assert(data.Schema.Label.HasValue); 300if (data.Schema.Label.Value.Type == BooleanDataViewType.Instance)