public ColumnInfo(string input, string output, OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutKind,
int maxNumTerms = ValueToKeyMappingEstimator.Defaults.MaxNumTerms, TermTransform.SortOrder sort = ValueToKeyMappingEstimator.Defaults.Sort,
string[] term = null)
: base(input, output, maxNumTerms, sort, term, true)
{
OutputKind = outputKind;
}
/// <summary>
/// Convert a text column into hash-based one-hot encoded vector.
/// </summary>
/// <param name="catalog">The transform catalog</param>
/// <param name="inputColumn">The input column</param>
/// <param name="outputColumn">The output column. If <c>null</c>, <paramref name="inputColumn"/> is used.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 30, inclusive.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
/// <param name="outputKind">The conversion mode.</param>
/// <returns></returns>
public static OneHotHashEncodingEstimator OneHotHashEncoding(this TransformsCatalog.CategoricalTransforms catalog,
string inputColumn,
string outputColumn = null,
int hashBits = OneHotHashEncodingEstimator.Defaults.HashBits,
int invertHash = OneHotHashEncodingEstimator.Defaults.InvertHash,
OneHotEncodingTransformer.OutputKind outputKind = OneHotEncodingTransformer.OutputKind.Ind)
=> new OneHotHashEncodingEstimator(CatalogUtils.GetEnvironment(catalog), inputColumn, outputColumn, hashBits, invertHash, outputKind);
/// <summary>
/// Instantiates a new instance of <see cref="OneHotHashEncodingEstimator"/>.
/// </summary>
/// <param name="env">Host Environment.</param>
/// <param name="outputColumnName">Name of the column resulting from the transformation of <paramref name="inputColumnName"/>.</param>
/// <param name="inputColumnName">Name of the column to transform.
/// If set to <see langword="null"/>, the value of the <paramref name="outputColumnName"/> will be used as source.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 30, inclusive.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
/// <param name="outputKind">The type of output expected.</param>
internal OneHotHashEncodingEstimator(IHostEnvironment env,
string outputColumnName,
string inputColumnName = null,
int hashBits = Defaults.HashBits,
int invertHash = Defaults.InvertHash,
OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutputKind)
: this(env, new ColumnInfo(outputColumnName, inputColumnName ?? outputColumnName, outputKind, hashBits, invertHash : invertHash))
{
}
/// A helper method to create <see cref="OneHotHashEncodingEstimator"/> for public facing API.
/// <param name="env">Host Environment.</param>
/// <param name="inputColumn">Name of the input column.</param>
/// <param name="outputColumn">Name of the output column. If this is null '<paramref name="inputColumn"/>' will be used.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 30, inclusive.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
/// <param name="outputKind">The type of output expected.</param>
public OneHotHashEncodingEstimator(IHostEnvironment env,
string inputColumn,
string outputColumn,
int hashBits = OneHotHashEncodingEstimator.Defaults.HashBits,
int invertHash = OneHotHashEncodingEstimator.Defaults.InvertHash,
OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutputKind)
: this(env, new ColumnInfo(inputColumn, outputColumn ?? inputColumn, outputKind, hashBits, invertHash : invertHash))
{
}
/// <summary>
/// Describes how the transformer handles one column pair.
/// </summary>
/// <param name="name">Name of the column resulting from the transformation of <paramref name="inputColumnName"/>.</param>
/// <param name="inputColumnName">Name of column to transform. If set to <see langword="null"/>, the value of the <paramref name="name"/> will be used as source.</param>
/// <param name="outputKind">Kind of output: bag, indicator vector etc.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 31, inclusive.</param>
/// <param name="seed">Hashing seed.</param>
/// <param name="ordered">Whether the position of each term should be included in the hash.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
public ColumnInfo(string name, string inputColumnName = null,
OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutputKind,
int hashBits = Defaults.HashBits,
uint seed = Defaults.Seed,
bool ordered = Defaults.Ordered,
int invertHash = Defaults.InvertHash)
{
HashInfo = new HashingEstimator.ColumnInfo(name, inputColumnName ?? name, hashBits, seed, ordered, invertHash);
OutputKind = outputKind;
}
/// <summary>
/// A helper method to create <see cref="OneHotHashEncodingTransformer"/>.
/// </summary>
/// <param name="env">Host Environment.</param>
/// <param name="input">Input <see cref="IDataView"/>. This is the output from previous transform or loader.</param>
/// <param name="name">Name of the output column.</param>
/// <param name="source">Name of the column to be transformed. If this is null '<paramref name="name"/>' will be used.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 30, inclusive.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
/// <param name="outputKind">The type of output expected.</param>
private static IDataView Create(IHostEnvironment env,
IDataView input,
string name,
string source = null,
int hashBits = OneHotHashEncodingEstimator.Defaults.HashBits,
int invertHash = OneHotHashEncodingEstimator.Defaults.InvertHash,
OneHotEncodingTransformer.OutputKind outputKind = OneHotHashEncodingEstimator.Defaults.OutputKind)
{
return(new OneHotHashEncodingEstimator(env, name, source, hashBits, invertHash, outputKind).Fit(input).Transform(input) as IDataView);
}
/// <summary>
/// Describes how the transformer handles one column pair.
/// </summary>
/// <param name="input">Name of input column.</param>
/// <param name="output">Name of output column.</param>
/// <param name="outputKind">Kind of output: bag, indicator vector etc.</param>
/// <param name="hashBits">Number of bits to hash into. Must be between 1 and 31, inclusive.</param>
/// <param name="seed">Hashing seed.</param>
/// <param name="ordered">Whether the position of each term should be included in the hash.</param>
/// <param name="invertHash">During hashing we constuct mappings between original values and the produced hash values.
/// Text representation of original values are stored in the slot names of the metadata for the new column.Hashing, as such, can map many initial values to one.
/// <paramref name="invertHash"/> specifies the upper bound of the number of distinct input values mapping to a hash that should be retained.
/// <value>0</value> does not retain any input values. <value>-1</value> retains all input values mapping to each hash.</param>
public ColumnInfo(string input, string output,
OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutputKind,
int hashBits = Defaults.HashBits,
uint seed = Defaults.Seed,
bool ordered = Defaults.Ordered,
int invertHash = Defaults.InvertHash)
{
HashInfo = new HashingTransformer.ColumnInfo(input, output, hashBits, seed, ordered, invertHash);
OutputKind = outputKind;
}
public OneHotEncodingEstimator(IHostEnvironment env, ColumnInfo[] columns,
string file = null, string termsColumn = null,
IComponentFactory <IMultiStreamSource, IDataLoader> loaderFactory = null)
{
Contracts.CheckValue(env, nameof(env));
_host = env.Register(nameof(OneHotEncodingEstimator));
_term = new ValueToKeyMappingEstimator(_host, columns, file, termsColumn, loaderFactory);
var binaryCols = new List <(string input, string output)>();
var cols = new List <(string input, string output, bool bag)>();
for (int i = 0; i < columns.Length; i++)
{
var column = columns[i];
OneHotEncodingTransformer.OutputKind kind = columns[i].OutputKind;
switch (kind)
{
default:
throw _host.ExceptUserArg(nameof(column.OutputKind));
case OneHotEncodingTransformer.OutputKind.Key:
continue;
case OneHotEncodingTransformer.OutputKind.Bin:
binaryCols.Add((column.Output, column.Output));
break;
case OneHotEncodingTransformer.OutputKind.Ind:
cols.Add((column.Output, column.Output, false));
break;
case OneHotEncodingTransformer.OutputKind.Bag:
cols.Add((column.Output, column.Output, true));
break;
}
}
IEstimator <ITransformer> toBinVector = null;
IEstimator <ITransformer> toVector = null;
if (binaryCols.Count > 0)
{
toBinVector = new KeyToBinaryVectorMappingEstimator(_host, binaryCols.Select(x => new KeyToBinaryVectorMappingTransformer.ColumnInfo(x.input, x.output)).ToArray());
}
if (cols.Count > 0)
{
toVector = new KeyToVectorMappingEstimator(_host, cols.Select(x => new KeyToVectorMappingTransformer.ColumnInfo(x.input, x.output, x.bag)).ToArray());
}
if (toBinVector != null && toVector != null)
{
_toSomething = toVector.Append(toBinVector);
}
else
{
if (toBinVector != null)
{
_toSomething = toBinVector;
}
else
{
_toSomething = toVector;
}
}
}
/// Initializes an instance of the <see cref="OneHotEncodingEstimator"/>.
/// <param name="env">Host Environment.</param>
/// <param name="inputColumn">Name of the column to be transformed.</param>
/// <param name="outputColumn">Name of the output column. If this is <c>null</c>, <paramref name="inputColumn"/> is used.</param>
/// <param name="outputKind">The type of output expected.</param>
public OneHotEncodingEstimator(IHostEnvironment env, string inputColumn,
string outputColumn = null, OneHotEncodingTransformer.OutputKind outputKind = Defaults.OutKind)
: this(env, new[] { new ColumnInfo(inputColumn, outputColumn ?? inputColumn, outputKind) })
{
}
internal OneHotHashEncodingEstimator(IHostEnvironment env, params ColumnInfo[] columns)
{
Contracts.CheckValue(env, nameof(env));
_host = env.Register(nameof(ValueToKeyMappingEstimator));
_hash = new HashingEstimator(_host, columns.Select(x => x.HashInfo).ToArray());
using (var ch = _host.Start(nameof(OneHotHashEncodingEstimator)))
{
var binaryCols = new List <(string outputColumnName, string inputColumnName)>();
var cols = new List <(string outputColumnName, string inputColumnName, bool bag)>();
for (int i = 0; i < columns.Length; i++)
{
var column = columns[i];
OneHotEncodingTransformer.OutputKind kind = columns[i].OutputKind;
switch (kind)
{
default:
throw _host.ExceptUserArg(nameof(column.OutputKind));
case OneHotEncodingTransformer.OutputKind.Key:
continue;
case OneHotEncodingTransformer.OutputKind.Bin:
if ((column.HashInfo.InvertHash) != 0)
{
ch.Warning("Invert hashing is being used with binary encoding.");
}
binaryCols.Add((column.HashInfo.Name, column.HashInfo.Name));
break;
case OneHotEncodingTransformer.OutputKind.Ind:
cols.Add((column.HashInfo.Name, column.HashInfo.Name, false));
break;
case OneHotEncodingTransformer.OutputKind.Bag:
cols.Add((column.HashInfo.Name, column.HashInfo.Name, true));
break;
}
}
IEstimator <ITransformer> toBinVector = null;
IEstimator <ITransformer> toVector = null;
if (binaryCols.Count > 0)
{
toBinVector = new KeyToBinaryVectorMappingEstimator(_host, binaryCols.Select(x => (x.outputColumnName, x.inputColumnName)).ToArray());
}
if (cols.Count > 0)
{
toVector = new KeyToVectorMappingEstimator(_host, cols.Select(x => new KeyToVectorMappingEstimator.ColumnInfo(x.outputColumnName, x.inputColumnName, x.bag)).ToArray());
}
if (toBinVector != null && toVector != null)
{
_toSomething = toVector.Append(toBinVector);
}
else
{
if (toBinVector != null)
{
_toSomething = toBinVector;
}
else
{
_toSomething = toVector;
}
}
}
}
/// <summary>
/// Convert a text column into one-hot encoded vector.
/// </summary>
/// <param name="catalog">The transform catalog</param>
/// <param name="inputColumn">The input column</param>
/// <param name="outputColumn">The output column. If <c>null</c>, <paramref name="inputColumn"/> is used.</param>
/// <param name="outputKind">The conversion mode.</param>
/// <returns></returns>
public static OneHotEncodingEstimator OneHotEncoding(this TransformsCatalog.CategoricalTransforms catalog,
string inputColumn,
string outputColumn = null,
OneHotEncodingTransformer.OutputKind outputKind = OneHotEncodingTransformer.OutputKind.Ind)
=> new OneHotEncodingEstimator(CatalogUtils.GetEnvironment(catalog), inputColumn, outputColumn, outputKind);
