public static BindingsImpl Create(ModelLoadContext ctx, ISchema input,
IHostEnvironment env, ISchemaBindableMapper bindable,
Func <ColumnType, bool> outputTypeMatches, Func <ColumnType, ISchemaBoundRowMapper, ColumnType> getPredColType)
{
Contracts.AssertValue(env);
env.AssertValue(ctx);
// *** Binary format ***
// <base info>
// int: id of the scores column kind (metadata output)
// int: id of the column used for deriving the predicted label column
string suffix;
var roles = LoadBaseInfo(ctx, out suffix);
string scoreKind = ctx.LoadNonEmptyString();
string scoreCol = ctx.LoadNonEmptyString();
var mapper = bindable.Bind(env, RoleMappedSchema.Create(input, roles));
var rowMapper = mapper as ISchemaBoundRowMapper;
env.CheckParam(rowMapper != null, nameof(bindable), "Bindable expected to be an " + nameof(ISchemaBindableMapper) + "!");
// Find the score column of the mapper.
int scoreColIndex;
env.CheckDecode(mapper.OutputSchema.TryGetColumnIndex(scoreCol, out scoreColIndex));
var scoreType = mapper.OutputSchema.GetColumnType(scoreColIndex);
env.CheckDecode(outputTypeMatches(scoreType));
var predColType = getPredColType(scoreType, rowMapper);
return(new BindingsImpl(input, rowMapper, suffix, scoreKind, false, scoreColIndex, predColType));
}
public BindingsImpl ApplyToSchema(ISchema input, ISchemaBindableMapper bindable, IHostEnvironment env)
{
Contracts.AssertValue(env);
env.AssertValue(input);
env.AssertValue(bindable);
string scoreCol = RowMapper.OutputSchema.GetColumnName(ScoreColumnIndex);
var schema = RoleMappedSchema.Create(input, RowMapper.GetInputColumnRoles());
// Checks compatibility of the predictor input types.
var mapper = bindable.Bind(env, schema);
var rowMapper = mapper as ISchemaBoundRowMapper;
env.CheckParam(rowMapper != null, nameof(bindable), "Mapper must implement ISchemaBoundRowMapper");
int mapperScoreColumn;
bool tmp = rowMapper.OutputSchema.TryGetColumnIndex(scoreCol, out mapperScoreColumn);
env.Check(tmp, "Mapper doesn't have expected score column");
return(new BindingsImpl(input, rowMapper, Suffix, ScoreColumnKind, true, mapperScoreColumn, PredColType));
}
/// <summary>
/// Create the bindings given the env, bindable, input schema, column roles, and column name suffix.
/// </summary>
private static Bindings Create(IHostEnvironment env, ISchemaBindableMapper bindable, ISchema input,
IEnumerable <KeyValuePair <RoleMappedSchema.ColumnRole, string> > roles, string suffix, bool user = true)
{
Contracts.AssertValue(env);
Contracts.AssertValue(bindable);
Contracts.AssertValue(input);
Contracts.AssertValue(roles);
Contracts.AssertValueOrNull(suffix);
var mapper = bindable.Bind(env, RoleMappedSchema.Create(input, roles));
// We don't actually depend on this invariant, but if this assert fires it means the bindable
// did the wrong thing.
Contracts.Assert(mapper.InputSchema.Schema == input);
var rowMapper = mapper as ISchemaBoundRowMapper;
Contracts.Check(rowMapper != null, "Predictor expected to be a RowMapper!");
return(Create(input, rowMapper, suffix, user));
}
/// <summary>
/// Creates a RoleMappedData from the given schema and role/column-name pairs.
/// This skips null or empty column-names.
/// </summary>
public static RoleMappedData Create(IDataView data, IEnumerable <KeyValuePair <RoleMappedSchema.ColumnRole, string> > roles)
{
Contracts.CheckValue(data, nameof(data));
Contracts.CheckValue(roles, nameof(roles));
return(new RoleMappedData(data, RoleMappedSchema.Create(data.Schema, roles)));
}
/// <summary>
/// Creates a RoleMappedData from the given data with no column role assignments.
/// </summary>
public static RoleMappedData Create(IDataView data)
{
Contracts.CheckValue(data, nameof(data));
return(new RoleMappedData(data, RoleMappedSchema.Create(data.Schema)));
}
/// <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 sub = Args.Loader;
IDataLoader trainPipe = null;
if (sub.IsGood())
{
// The loader is overridden from the command line.
pipe = sub.CreateInstance(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 = RoleMappedSchema.Create(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.
}
}
}