private IDataLoader LoadTransformChain(IDataLoader srcData)
{
Host.Assert(!string.IsNullOrWhiteSpace(Args.InputModelFile));
using (var file = Host.OpenInputFile(Args.InputModelFile))
using (var strm = file.OpenReadStream())
using (var rep = RepositoryReader.Open(strm, Host))
using (var pipeLoaderEntry = rep.OpenEntry(ModelFileUtils.DirDataLoaderModel, ModelLoadContext.ModelStreamName))
using (var ctx = new ModelLoadContext(rep, pipeLoaderEntry, ModelFileUtils.DirDataLoaderModel))
return(CompositeDataLoader.Create(Host, ctx, srcData, x => true));
}
private void RunCore(IChannel ch)
{
Host.AssertValue(ch);
ch.Trace("Creating loader");
LoadModelObjects(ch, true, out var predictor, true, out var trainSchema, out var loader);
ch.AssertValue(predictor);
ch.AssertValueOrNull(trainSchema);
ch.AssertValue(loader);
ch.Trace("Creating pipeline");
var scorer = Args.Scorer;
ch.Assert(scorer == null || scorer is ICommandLineComponentFactory, "ScoreCommand should only be used from the command line.");
var bindable = ScoreUtils.GetSchemaBindableMapper(Host, predictor, scorerFactorySettings: scorer as ICommandLineComponentFactory);
ch.AssertValue(bindable);
// REVIEW: We probably ought to prefer role mappings from the training schema.
string feat = TrainUtils.MatchNameOrDefaultOrNull(ch, loader.Schema,
nameof(Args.FeatureColumn), Args.FeatureColumn, DefaultColumnNames.Features);
string group = TrainUtils.MatchNameOrDefaultOrNull(ch, loader.Schema,
nameof(Args.GroupColumn), Args.GroupColumn, DefaultColumnNames.GroupId);
var customCols = TrainUtils.CheckAndGenerateCustomColumns(ch, Args.CustomColumn);
var schema = new RoleMappedSchema(loader.Schema, label: null, feature: feat, group: group, custom: customCols, opt: true);
var mapper = bindable.Bind(Host, schema);
if (scorer == null)
{
scorer = ScoreUtils.GetScorerComponent(Host, mapper);
}
loader = CompositeDataLoader.ApplyTransform(Host, loader, "Scorer", scorer.ToString(),
(env, view) => scorer.CreateComponent(env, view, mapper, trainSchema));
loader = CompositeDataLoader.Create(Host, loader, Args.PostTransform);
if (!string.IsNullOrWhiteSpace(Args.OutputModelFile))
{
ch.Trace("Saving the data pipe");
SaveLoader(loader, Args.OutputModelFile);
}
ch.Trace("Creating saver");
IDataSaver writer;
if (Args.Saver == null)
{
var ext = Path.GetExtension(Args.OutputDataFile);
var isText = ext == ".txt" || ext == ".tlc";
if (isText)
{
writer = new TextSaver(Host, new TextSaver.Arguments());
}
else
{
writer = new BinarySaver(Host, new BinarySaver.Arguments());
}
}
else
{
writer = Args.Saver.CreateComponent(Host);
}
ch.Assert(writer != null);
var outputIsBinary = writer is BinaryWriter;
bool outputAllColumns =
Args.OutputAllColumns == true ||
(Args.OutputAllColumns == null && Utils.Size(Args.OutputColumn) == 0 && outputIsBinary);
bool outputNamesAndLabels =
Args.OutputAllColumns == true || Utils.Size(Args.OutputColumn) == 0;
if (Args.OutputAllColumns == true && Utils.Size(Args.OutputColumn) != 0)
{
ch.Warning(nameof(Args.OutputAllColumns) + "=+ always writes all columns irrespective of " + nameof(Args.OutputColumn) + " specified.");
}
if (!outputAllColumns && Utils.Size(Args.OutputColumn) != 0)
{
foreach (var outCol in Args.OutputColumn)
{
if (!loader.Schema.TryGetColumnIndex(outCol, out int dummyColIndex))
{
throw ch.ExceptUserArg(nameof(Arguments.OutputColumn), "Column '{0}' not found.", outCol);
}
}
}
uint maxScoreId = 0;
if (!outputAllColumns)
{
maxScoreId = loader.Schema.GetMaxMetadataKind(out int colMax, MetadataUtils.Kinds.ScoreColumnSetId);
}
ch.Assert(outputAllColumns || maxScoreId > 0); // score set IDs are one-based
var cols = new List <int>();
for (int i = 0; i < loader.Schema.Count; i++)
{
if (!Args.KeepHidden && loader.Schema.IsHidden(i))
{
continue;
}
if (!(outputAllColumns || ShouldAddColumn(loader.Schema, i, maxScoreId, outputNamesAndLabels)))
{
continue;
}
var type = loader.Schema.GetColumnType(i);
if (writer.IsColumnSavable(type))
{
cols.Add(i);
}
else
{
ch.Warning("The column '{0}' will not be written as it has unsavable column type.",
loader.Schema.GetColumnName(i));
}
}
ch.Check(cols.Count > 0, "No valid columns to save");
ch.Trace("Scoring and saving data");
using (var file = Host.CreateOutputFile(Args.OutputDataFile))
using (var stream = file.CreateWriteStream())
writer.SaveData(stream, loader, cols.ToArray());
}
private void RunCore(IChannel ch, string cmd)
{
Host.AssertValue(ch);
IPredictor inputPredictor = null;
if (Args.ContinueTrain && !TrainUtils.TryLoadPredictor(ch, Host, Args.InputModelFile, out inputPredictor))
{
ch.Warning("No input model file specified or model file did not contain a predictor. The model state cannot be initialized.");
}
ch.Trace("Constructing data pipeline");
IDataLoader loader = CreateRawLoader();
// If the per-instance results are requested and there is no name column, add a GenerateNumberTransform.
var preXf = Args.PreTransform;
if (!string.IsNullOrEmpty(Args.OutputDataFile))
{
string name = TrainUtils.MatchNameOrDefaultOrNull(ch, loader.Schema, nameof(Args.NameColumn), Args.NameColumn, DefaultColumnNames.Name);
if (name == null)
{
var args = new GenerateNumberTransform.Arguments();
args.Column = new[] { new GenerateNumberTransform.Column()
{
Name = DefaultColumnNames.Name
}, };
args.UseCounter = true;
var options = CmdParser.GetSettings(ch, args, new GenerateNumberTransform.Arguments());
preXf = preXf.Concat(
new[]
{
new KeyValuePair <string, SubComponent <IDataTransform, SignatureDataTransform> >(
"", new SubComponent <IDataTransform, SignatureDataTransform>(
GenerateNumberTransform.LoadName, options))
}).ToArray();
}
}
loader = CompositeDataLoader.Create(Host, loader, preXf);
ch.Trace("Binding label and features columns");
IDataView pipe = loader;
var stratificationColumn = GetSplitColumn(ch, loader, ref pipe);
var scorer = Args.Scorer;
var evaluator = Args.Evaluator;
Func <IDataView> validDataCreator = null;
if (Args.ValidationFile != null)
{
validDataCreator =
() =>
{
// Fork the command.
var impl = new CrossValidationCommand(this);
return(impl.CreateRawLoader(dataFile: Args.ValidationFile));
};
}
FoldHelper fold = new FoldHelper(Host, RegistrationName, pipe, stratificationColumn,
Args, CreateRoleMappedData, ApplyAllTransformsToData, scorer, evaluator,
validDataCreator, ApplyAllTransformsToData, inputPredictor, cmd, loader, !string.IsNullOrEmpty(Args.OutputDataFile));
var tasks = fold.GetCrossValidationTasks();
if (!evaluator.IsGood())
{
evaluator = EvaluateUtils.GetEvaluatorType(ch, tasks[0].Result.ScoreSchema);
}
var eval = evaluator.CreateInstance(Host);
// Print confusion matrix and fold results for each fold.
for (int i = 0; i < tasks.Length; i++)
{
var dict = tasks[i].Result.Metrics;
MetricWriter.PrintWarnings(ch, dict);
eval.PrintFoldResults(ch, dict);
}
// Print the overall results.
if (!TryGetOverallMetrics(tasks.Select(t => t.Result.Metrics).ToArray(), out var overallList))
{
throw ch.Except("No overall metrics found");
}
var overall = eval.GetOverallResults(overallList.ToArray());
MetricWriter.PrintOverallMetrics(Host, ch, Args.SummaryFilename, overall, Args.NumFolds);
eval.PrintAdditionalMetrics(ch, tasks.Select(t => t.Result.Metrics).ToArray());
Dictionary <string, IDataView>[] metricValues = tasks.Select(t => t.Result.Metrics).ToArray();
SendTelemetryMetric(metricValues);
// Save the per-instance results.
if (!string.IsNullOrWhiteSpace(Args.OutputDataFile))
{
var perInstance = EvaluateUtils.ConcatenatePerInstanceDataViews(Host, eval, Args.CollateMetrics,
Args.OutputExampleFoldIndex, tasks.Select(t => t.Result.PerInstanceResults).ToArray(), out var variableSizeVectorColumnNames);
if (variableSizeVectorColumnNames.Length > 0)
{
ch.Warning("Detected columns of variable length: {0}. Consider setting collateMetrics- for meaningful per-Folds results.",
string.Join(", ", variableSizeVectorColumnNames));
}
if (Args.CollateMetrics)
{
ch.Assert(perInstance.Length == 1);
MetricWriter.SavePerInstance(Host, ch, Args.OutputDataFile, perInstance[0]);
}
else
{
int i = 0;
foreach (var idv in perInstance)
{
MetricWriter.SavePerInstance(Host, ch, ConstructPerFoldName(Args.OutputDataFile, i), idv);
i++;
}
}
}
}
private IDataLoader CreateTransformChain(IDataLoader loader)
{
return(CompositeDataLoader.Create(Host, loader, Args.Transform));
}
/// <summary>
/// Loads multiple artifacts of interest from the input model file, given the context
/// established by the command line arguments.
/// </summary>
/// <param name="ch">The channel to which to provide output.</param>
/// <param name="wantPredictor">Whether we want a predictor from the model file. If
/// <c>false</c> we will not even attempt to load a predictor. If <c>null</c> we will
/// load the predictor, if present. If <c>true</c> we will load the predictor, or fail
/// noisily if we cannot.</param>
/// <param name="predictor">The predictor in the model, or <c>null</c> if
/// <paramref name="wantPredictor"/> was false, or <paramref name="wantPredictor"/> was
/// <c>null</c> and no predictor was present.</param>
/// <param name="wantTrainSchema">Whether we want the training schema. Unlike
/// <paramref name="wantPredictor"/>, this has no "hard fail if not present" option. If
/// this is <c>true</c>, it is still possible for <paramref name="trainSchema"/> to remain
/// <c>null</c> if there were no role mappings, or pipeline.</param>
/// <param name="trainSchema">The training schema if <paramref name="wantTrainSchema"/>
/// is true, and there were role mappings stored in the model.</param>
/// <param name="pipe">The data pipe constructed from the combination of the
/// model and command line arguments.</param>
protected void LoadModelObjects(
IChannel ch,
bool?wantPredictor, out IPredictor predictor,
bool wantTrainSchema, out RoleMappedSchema trainSchema,
out IDataLoader pipe)
{
// First handle the case where there is no input model file.
// Everything must come from the command line.
using (var file = Host.OpenInputFile(Args.InputModelFile))
using (var strm = file.OpenReadStream())
using (var rep = RepositoryReader.Open(strm, Host))
{
// First consider loading the predictor.
if (wantPredictor == false)
{
predictor = null;
}
else
{
ch.Trace("Loading predictor");
predictor = ModelFileUtils.LoadPredictorOrNull(Host, rep);
if (wantPredictor == true)
{
Host.Check(predictor != null, "Could not load predictor from model file");
}
}
// Next create the loader.
var loaderFactory = Args.Loader;
IDataLoader trainPipe = null;
if (loaderFactory != null)
{
// The loader is overridden from the command line.
pipe = loaderFactory.CreateComponent(Host, new MultiFileSource(Args.DataFile));
if (Args.LoadTransforms == true)
{
Host.CheckUserArg(!string.IsNullOrWhiteSpace(Args.InputModelFile), nameof(Args.InputModelFile));
pipe = LoadTransformChain(pipe);
}
}
else
{
var loadTrans = Args.LoadTransforms ?? true;
pipe = LoadLoader(rep, Args.DataFile, loadTrans);
if (loadTrans)
{
trainPipe = pipe;
}
}
if (Utils.Size(Args.Transform) > 0)
{
pipe = CompositeDataLoader.Create(Host, pipe, Args.Transform);
}
// Next consider loading the training data's role mapped schema.
trainSchema = null;
if (wantTrainSchema)
{
// First try to get the role mappings.
var trainRoleMappings = ModelFileUtils.LoadRoleMappingsOrNull(Host, rep);
if (trainRoleMappings != null)
{
// Next create the training schema. In the event that the loaded pipeline happens
// to be the training pipe, we can just use that. If it differs, then we need to
// load the full pipeline from the model, relying upon the fact that all loaders
// can be loaded with no data at all, to get their schemas.
if (trainPipe == null)
{
trainPipe = ModelFileUtils.LoadLoader(Host, rep, new MultiFileSource(null), loadTransforms: true);
}
trainSchema = new RoleMappedSchema(trainPipe.Schema, trainRoleMappings);
}
// If the role mappings are null, an alternative would be to fail. However the idea
// is that the scorer should always still succeed, although perhaps with reduced
// functionality, even when the training schema is null, since not all versions of
// TLC models will have the role mappings preserved, I believe. And, we do want to
// maintain backwards compatibility.
}
}
}
private void RunCore(IChannel ch, string cmd)
{
Host.AssertValue(ch);
IPredictor inputPredictor = null;
if (Args.ContinueTrain && !TrainUtils.TryLoadPredictor(ch, Host, Args.InputModelFile, out inputPredictor))
{
ch.Warning("No input model file specified or model file did not contain a predictor. The model state cannot be initialized.");
}
ch.Trace("Constructing data pipeline");
IDataLoader loader = CreateRawLoader();
// If the per-instance results are requested and there is no name column, add a GenerateNumberTransform.
var preXf = Args.PreTransform;
if (!string.IsNullOrEmpty(Args.OutputDataFile))
{
string name = TrainUtils.MatchNameOrDefaultOrNull(ch, loader.Schema, nameof(Args.NameColumn), Args.NameColumn, DefaultColumnNames.Name);
if (name == null)
{
var args = new GenerateNumberTransform.Arguments();
args.Column = new[] { new GenerateNumberTransform.Column()
{
Name = DefaultColumnNames.Name
}, };
args.UseCounter = true;
var options = CmdParser.GetSettings(ch, args, new GenerateNumberTransform.Arguments());
preXf = preXf.Concat(
new[]
{
new KeyValuePair <string, SubComponent <IDataTransform, SignatureDataTransform> >(
"", new SubComponent <IDataTransform, SignatureDataTransform>(
GenerateNumberTransform.LoadName, options))
}).ToArray();
}
}
loader = CompositeDataLoader.Create(Host, loader, preXf);
ch.Trace("Binding label and features columns");
IDataView pipe = loader;
var stratificationColumn = GetSplitColumn(ch, loader, ref pipe);
var scorer = Args.Scorer;
var evaluator = Args.Evaluator;
Func <IDataView> validDataCreator = null;
if (Args.ValidationFile != null)
{
validDataCreator =
() =>
{
// Fork the command.
var impl = new CrossValidationCommand(this);
return(impl.CreateRawLoader(dataFile: Args.ValidationFile));
};
}
FoldHelper fold = new FoldHelper(Host, RegistrationName, pipe, stratificationColumn,
Args, CreateRoleMappedData, ApplyAllTransformsToData, scorer, evaluator,
validDataCreator, ApplyAllTransformsToData, inputPredictor, cmd, loader, !string.IsNullOrEmpty(Args.OutputDataFile));
var tasks = fold.GetCrossValidationTasks();
if (!evaluator.IsGood())
{
evaluator = EvaluateUtils.GetEvaluatorType(ch, tasks[0].Result.ScoreSchema);
}
var eval = evaluator.CreateInstance(Host);
// Print confusion matrix and fold results for each fold.
for (int i = 0; i < tasks.Length; i++)
{
var dict = tasks[i].Result.Metrics;
MetricWriter.PrintWarnings(ch, dict);
eval.PrintFoldResults(ch, dict);
}
// Print the overall results.
eval.PrintOverallResults(ch, Args.SummaryFilename, tasks.Select(t => t.Result.Metrics).ToArray());
Dictionary <string, IDataView>[] metricValues = tasks.Select(t => t.Result.Metrics).ToArray();
SendTelemetryMetric(metricValues);
// Save the per-instance results.
if (!string.IsNullOrWhiteSpace(Args.OutputDataFile))
{
Func <Task <FoldHelper.FoldResult>, int, IDataView> getPerInstance =
(task, i) =>
{
if (!Args.OutputExampleFoldIndex)
{
return(task.Result.PerInstanceResults);
}
// If the fold index is requested, add a column containing it. We use the first column in the data view
// as an input column to the LambdaColumnMapper, because it must have an input.
var inputColName = task.Result.PerInstanceResults.Schema.GetColumnName(0);
var inputColType = task.Result.PerInstanceResults.Schema.GetColumnType(0);
return(Utils.MarshalInvoke(EvaluateUtils.AddKeyColumn <int>, inputColType.RawType, Host,
task.Result.PerInstanceResults, inputColName, MetricKinds.ColumnNames.FoldIndex,
inputColType, Args.NumFolds, i + 1, "FoldIndex", default(ValueGetter <VBuffer <DvText> >)));
};
var foldDataViews = tasks.Select(getPerInstance).ToArray();
if (Args.CollateMetrics)
{
var perInst = AppendPerInstanceDataViews(foldDataViews, ch);
MetricWriter.SavePerInstance(Host, ch, Args.OutputDataFile, perInst);
}
else
{
int i = 0;
foreach (var idv in foldDataViews)
{
MetricWriter.SavePerInstance(Host, ch, ConstructPerFoldName(Args.OutputDataFile, i), idv);
i++;
}
}
}
}
// Returns true if a normalizer was added.
public static bool AddNormalizerIfNeeded(IHostEnvironment env, IChannel ch, ITrainer trainer, ref IDataView view, string featureColumn, NormalizeOption autoNorm)
{
Contracts.CheckValue(env, nameof(env));
env.CheckValue(ch, nameof(ch));
ch.CheckValue(trainer, nameof(trainer));
ch.CheckValue(view, nameof(view));
ch.CheckValueOrNull(featureColumn);
ch.CheckUserArg(Enum.IsDefined(typeof(NormalizeOption), autoNorm), nameof(TrainCommand.Arguments.NormalizeFeatures),
"Normalize option is invalid. Specify one of 'norm=No', 'norm=Warn', 'norm=Auto', or 'norm=Yes'.");
if (autoNorm == NormalizeOption.No)
{
ch.Info("Not adding a normalizer.");
return(false);
}
if (string.IsNullOrEmpty(featureColumn))
{
return(false);
}
int featCol;
var schema = view.Schema;
if (schema.TryGetColumnIndex(featureColumn, out featCol))
{
if (autoNorm != NormalizeOption.Yes)
{
var nn = trainer as ITrainerEx;
DvBool isNormalized = DvBool.False;
if (nn == null || !nn.NeedNormalization ||
(schema.TryGetMetadata(BoolType.Instance, MetadataUtils.Kinds.IsNormalized, featCol, ref isNormalized) &&
isNormalized.IsTrue))
{
ch.Info("Not adding a normalizer.");
return(false);
}
if (autoNorm == NormalizeOption.Warn)
{
ch.Warning("A normalizer is needed for this trainer. Either add a normalizing transform or use the 'norm=Auto', 'norm=Yes' or 'norm=No' options.");
return(false);
}
}
ch.Info("Automatically adding a MinMax normalization transform, use 'norm=Warn' or 'norm=No' to turn this behavior off.");
// Quote the feature column name
string quotedFeatureColumnName = featureColumn;
StringBuilder sb = new StringBuilder();
if (CmdQuoter.QuoteValue(quotedFeatureColumnName, sb))
{
quotedFeatureColumnName = sb.ToString();
}
var component = new SubComponent <IDataTransform, SignatureDataTransform>("MinMax", string.Format("col={{ name={0} source={0} }}", quotedFeatureColumnName));
var loader = view as IDataLoader;
if (loader != null)
{
view = CompositeDataLoader.Create(env, loader,
new KeyValuePair <string, SubComponent <IDataTransform, SignatureDataTransform> >(null, component));
}
else
{
view = component.CreateInstance(env, view);
}
return(true);
}
return(false);
}