mlpack 4.6.2__cp313-cp313-win_amd64.whl → 4.7.0__cp313-cp313-win_amd64.whl

mlpack/include/mlpack/core/data/save_sparse.hpp

@@ -0,0 +1,44 @@
+/**
+ * @file core/data/save_sparse.hpp
+ * @author Ryan Curtin
+ * @author Omar Shrit
+ *
+ * Internal implementation of sparse matrix save function.
+ *
+ * mlpack is free software; you may redistribute it and/or modify it under the
+ * terms of the 3-clause BSD license.  You should have received a copy of the
+ * 3-clause BSD license along with mlpack.  If not, see
+ * http://www.opensource.org/licenses/BSD-3-Clause for more information.
+ */
+#ifndef MLPACK_CORE_DATA_SAVE_SPARSE_HPP
+#define MLPACK_CORE_DATA_SAVE_SPARSE_HPP
+
+#include "save_matrix.hpp"
+
+namespace mlpack {
+
+// Save a Sparse Matrix
+template<typename eT>
+bool SaveSparse(const arma::SpMat<eT>& matrix,
+                TextOptions& opts,
+                const std::string& filename,
+                std::fstream& stream)
+{
+  bool success = false;
+  arma::SpMat<eT> tmp;
+
+  // Transpose the matrix.
+  if (!opts.NoTranspose())
+  {
+    arma::SpMat<eT> tmp = trans(matrix);
+    success = SaveMatrix(tmp, opts, filename, stream);
+  }
+  else
+    success = SaveMatrix(matrix, opts, filename, stream);
+
+  return success;
+}
+
+} // namespace mlpack
+
+#endif

mlpack/include/mlpack/core/data/scaler_methods/max_abs_scaler.hpp

@@ -15,7 +15,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple MaxAbs Scaler class.
@@ -116,7 +115,6 @@ class MaxAbsScaler
   arma::vec scale;
 }; // class MaxAbsScaler
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/scaler_methods/mean_normalization.hpp

@@ -15,7 +15,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple Mean Normalization class
@@ -54,9 +53,9 @@ class MeanNormalization
   template<typename MatType>
   void Fit(const MatType& input)
   {
-    itemMean = arma::mean(input, 1);
+    itemMean = mean(input, 1);
     itemMin = min(input, 1);
-    itemMax = arma::max(input, 1);
+    itemMax = max(input, 1);
     scale = itemMax - itemMin;
     // Handling zeros in scale vector.
     scale.for_each([](arma::vec::elem_type& val) { val =
@@ -123,7 +122,6 @@ class MeanNormalization
   arma::vec scale;
 }; // class MeanNormalization
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/scaler_methods/min_max_scaler.hpp

@@ -15,7 +15,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple MinMax Scaler class
@@ -153,7 +152,6 @@ class MinMaxScaler
   arma::vec scalerowmin;
 }; // class MinMaxScaler
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/scaler_methods/pca_whitening.hpp

@@ -16,7 +16,6 @@
 #include <mlpack/core/math/ccov.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple PCAWhitening class.
@@ -69,7 +68,7 @@ class PCAWhitening
   template<typename MatType>
   void Fit(const MatType& input)
   {
-    itemMean = arma::mean(input, 1);
+    itemMean = mean(input, 1);
     // Get eigenvectors and eigenvalues of covariance of input matrix.
     eig_sym(eigenValues, eigenVectors, ColumnCovariance(
         input.each_col() - itemMean));
@@ -139,7 +138,6 @@ class PCAWhitening
   arma::vec eigenValues;
 }; // class PCAWhitening
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/scaler_methods/standard_scaler.hpp

@@ -15,7 +15,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple Standard Scaler class
@@ -55,8 +54,8 @@ class StandardScaler
   template<typename MatType>
   void Fit(const MatType& input)
   {
-    itemMean = arma::mean(input, 1);
-    itemStdDev = arma::stddev(input, 1, 1);
+    itemMean = mean(input, 1);
+    itemStdDev = stddev(input, 1, 1);
     // Handle zeros in scale vector.
     itemStdDev.for_each([](arma::vec::elem_type& val) { val =
         (val == 0) ? 1 : val; });
@@ -112,7 +111,6 @@ class StandardScaler
   arma::vec itemStdDev;
 }; // class StandardScaler
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/scaler_methods/zca_whitening.hpp

@@ -16,7 +16,6 @@
 #include <mlpack/core/data/scaler_methods/pca_whitening.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * A simple ZCAWhitening class.
@@ -111,7 +110,6 @@ class ZCAWhitening
   PCAWhitening pca;
 }; // class ZCAWhitening
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/split_data.hpp

@@ -16,7 +16,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * This helper function splits any `input` data into training and testing parts.
@@ -191,9 +190,9 @@ void StratifiedSplit(const arma::Mat<T>& input,
   const bool typeCheck = (arma::is_Row<LabelsType>::value)
       || (arma::is_Col<LabelsType>::value);
   if (!typeCheck)
-    throw std::runtime_error("data::Split(): when stratified sampling is done, "
+    throw std::runtime_error("Split(): when stratified sampling is done, "
         "labels must have type `arma::Row<>`!");
-  util::CheckSameSizes(input, inputLabel, "data::Split()");
+  util::CheckSameSizes(input, inputLabel, "Split()");
   size_t trainIdx = 0;
   size_t testIdx = 0;
   size_t trainSize = 0;
@@ -309,7 +308,7 @@ void Split(const MatType& input,
            const double testRatio,
            const bool shuffleData = true)
 {
-  util::CheckSameSizes(input, inputLabel, "data::Split()");
+  util::CheckSameSizes(input, inputLabel, "Split()");
   if (shuffleData)
   {
     arma::uvec order = arma::shuffle(arma::linspace<arma::uvec>(0,
@@ -422,8 +421,8 @@ void Split(const MatType& input,
            const double testRatio,
            const bool shuffleData = true)
 {
-  util::CheckSameSizes(input, inputLabels, "data::Split()");
-  util::CheckSameSizes(input, inputWeights, "data::Split()");
+  util::CheckSameSizes(input, inputLabels, "Split()");
+  util::CheckSameSizes(input, inputWeights, "Split()");
   if (shuffleData)
   {
     arma::uvec order = arma::shuffle(arma::linspace<arma::uvec>(0,
@@ -625,7 +624,7 @@ void Split(const FieldType& input,
            const double testRatio,
            const bool shuffleData = true)
 {
-  util::CheckSameSizes(input, inputLabel, "data::Split()");
+  util::CheckSameSizes(input, inputLabel, "Split()");
   if (shuffleData)
   {
     arma::uvec order = arma::shuffle(arma::linspace<arma::uvec>(0,
@@ -733,7 +732,6 @@ Split(const FieldType& input,
                          std::move(testData));
 }
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_algorithms.hpp

@@ -14,7 +14,6 @@
 #define MLPACK_CORE_DATA_STRING_ALGORITHMS_HPP
 
 namespace mlpack {
-namespace data {
 
 /**
  * Trim off characters from start and end of
@@ -122,7 +121,6 @@ inline std::vector<std::string> Tokenize(
   return tokens;
 }
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding.hpp

@@ -19,7 +19,6 @@
 #include <vector>
 
 namespace mlpack {
-namespace data {
 
 /**
  * The class translates a set of strings into numbers using various encoding
@@ -208,7 +207,6 @@ class StringEncoding
   DictionaryType dictionary;
 };
 
-} // namespace data
 } // namespace mlpack
 
 // Include implementation.

mlpack/include/mlpack/core/data/string_encoding_dictionary.hpp

@@ -21,7 +21,6 @@
 #include <unordered_map>
 
 namespace mlpack {
-namespace data {
 
 /**
  * This class provides a dictionary interface for the purpose of string
@@ -343,7 +342,6 @@ class StringEncodingDictionary<int>
   size_t size;
 };
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding_impl.hpp

@@ -18,7 +18,6 @@
 #include <type_traits>
 
 namespace mlpack {
-namespace data {
 
 template<typename EncodingPolicyType, typename DictionaryType>
 template<typename ... ArgTypes>
@@ -205,7 +204,6 @@ void StringEncoding<EncodingPolicyType, DictionaryType>::serialize(
   ar(CEREAL_NVP(dictionary));
 }
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding_policies/bag_of_words_encoding_policy.hpp

@@ -18,7 +18,6 @@
 #include <mlpack/core/data/string_encoding.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * Definition of the BagOfWordsEncodingPolicy class.
@@ -165,7 +164,6 @@ class BagOfWordsEncodingPolicy
 template<typename TokenType>
 using BagOfWordsEncoding = StringEncoding<BagOfWordsEncodingPolicy,
                                           StringEncodingDictionary<TokenType>>;
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding_policies/dictionary_encoding_policy.hpp

@@ -18,7 +18,6 @@
 #include <mlpack/core/data/string_encoding.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * DicitonaryEnocdingPolicy is used as a helper class for StringEncoding.
@@ -144,7 +143,6 @@ struct StringEncodingPolicyTraits<DictionaryEncodingPolicy>
 template<typename TokenType>
 using DictionaryEncoding = StringEncoding<DictionaryEncodingPolicy,
                                           StringEncodingDictionary<TokenType>>;
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding_policies/policy_traits.hpp

@@ -17,7 +17,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * This is a template struct that provides some information about various
@@ -33,7 +32,6 @@ struct StringEncodingPolicyTraits
   static const bool onePassEncoding = false;
 };
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/string_encoding_policies/tf_idf_encoding_policy.hpp

@@ -18,7 +18,6 @@
 #include <mlpack/core/data/string_encoding.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * Definition of the TfIdfEncodingPolicy class. TfIdfEncodingPolicy is used
@@ -343,7 +342,6 @@ class TfIdfEncodingPolicy
 template<typename TokenType>
 using TfIdfEncoding = StringEncoding<TfIdfEncodingPolicy,
                                      StringEncodingDictionary<TokenType>>;
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/text_options.hpp

@@ -19,46 +19,47 @@
 #include "matrix_options.hpp"
 
 namespace mlpack {
-namespace data {
 
 class TextOptions : public MatrixOptionsBase<TextOptions>
 {
  public:
   // TODO: pass through noTranspose option?
-  TextOptions(bool hasHeaders = defaultHasHeaders,
-              bool semicolon = defaultSemicolon,
-              bool missingToNan = defaultMissingToNan,
-              bool categorical = defaultCategorical,
-              bool missingPolicy = defaultMissingPolicy) :
+  TextOptions(std::optional<bool> hasHeaders = std::nullopt,
+              std::optional<bool> semicolon = std::nullopt,
+              std::optional<bool> missingToNan = std::nullopt,
+              std::optional<bool> categorical = std::nullopt) :
       MatrixOptionsBase<TextOptions>(),
       hasHeaders(hasHeaders),
       semicolon(semicolon),
       missingToNan(missingToNan),
-      categorical(categorical),
-      missingPolicy(missingPolicy)
+      categorical(categorical)
   {
     // Do Nothing.
   }
 
-  explicit TextOptions(const TextOptions& opts) :
+  //
+  // Handling for copy and move operations on other TextOptions objects.
+  //
+
+  TextOptions(const DataOptionsBase<MatrixOptionsBase<TextOptions>>& opts) :
       MatrixOptionsBase<TextOptions>()
   {
     // Delegate to copy operator.
     *this = opts;
   }
 
-  explicit TextOptions(TextOptions&& opts) :
+  TextOptions(DataOptionsBase<MatrixOptionsBase<TextOptions>>&& opts) :
       MatrixOptionsBase<TextOptions>()
   {
     // Delegate to move operator.
     *this = std::move(opts);
   }
 
-  // Inherit base class constructors.
-  using MatrixOptionsBase<TextOptions>::MatrixOptionsBase;
-
-  TextOptions& operator=(const TextOptions& other)
+  TextOptions& operator=(
+      const DataOptionsBase<MatrixOptionsBase<TextOptions>>& otherIn)
   {
+    const TextOptions& other = static_cast<const TextOptions&>(otherIn);
+
     if (&other == this)
       return *this;
 
@@ -70,12 +71,9 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
       missingToNan = *other.missingToNan;
     if (other.categorical.has_value())
       categorical = *other.categorical;
-    if (other.missingPolicy.has_value())
-      missingPolicy = *other.missingPolicy;
 
     headers = other.headers;
     datasetInfo = other.datasetInfo;
-    datasetMissingPolicy = other.datasetMissingPolicy;
 
     // Copy base members.
     MatrixOptionsBase<TextOptions>::operator=(other);
@@ -83,8 +81,11 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     return *this;
   }
 
-  TextOptions& operator=(TextOptions&& other)
+  TextOptions& operator=(
+      DataOptionsBase<MatrixOptionsBase<TextOptions>>&& otherIn)
   {
+    TextOptions&& other = static_cast<TextOptions&&>(otherIn);
+
     if (&other == this)
       return *this;
 
@@ -92,11 +93,9 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     semicolon = std::move(other.semicolon);
     missingToNan = std::move(other.missingToNan);
     categorical = std::move(other.categorical);
-    missingPolicy = std::move(other.missingPolicy);
 
     headers = std::move(other.headers);
     datasetInfo = std::move(other.datasetInfo);
-    datasetMissingPolicy = std::move(other.datasetMissingPolicy);
 
     // Move base members.
     MatrixOptionsBase<TextOptions>::operator=(std::move(other));
@@ -104,6 +103,55 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     return *this;
   }
 
+  //
+  // Handling for copy and move operations on other DataOptionsBase types.
+  //
+
+  // Conversions must be explicit.
+  template<typename Derived2>
+  explicit TextOptions(const DataOptionsBase<Derived2>& other) :
+      MatrixOptionsBase<TextOptions>(other) { }
+
+  template<typename Derived2>
+  explicit TextOptions(DataOptionsBase<Derived2>&& other) :
+      MatrixOptionsBase<TextOptions>(std::move(other)) { }
+
+  template<typename Derived2>
+  TextOptions& operator=(const DataOptionsBase<Derived2>& other)
+  {
+    return static_cast<TextOptions&>(
+        MatrixOptionsBase<TextOptions>::operator=(other));
+  }
+
+  template<typename Derived2>
+  TextOptions& operator=(DataOptionsBase<Derived2>&& other)
+  {
+    return static_cast<TextOptions&>(
+        MatrixOptionsBase<TextOptions>::operator=(std::move(other)));
+  }
+
+  void Combine(const TextOptions& other)
+  {
+    // Combine all boolean options.
+    hasHeaders =
+        DataOptionsBase<MatrixOptionsBase<TextOptions>>::CombineBooleanOption(
+        hasHeaders, other.hasHeaders, "HasHeaders()");
+    missingToNan =
+        DataOptionsBase<MatrixOptionsBase<TextOptions>>::CombineBooleanOption(
+        missingToNan, other.missingToNan, "MissingToNan()");
+    categorical =
+        DataOptionsBase<MatrixOptionsBase<TextOptions>>::CombineBooleanOption(
+        categorical, other.categorical, "Categorical()");
+    semicolon =
+        DataOptionsBase<MatrixOptionsBase<TextOptions>>::CombineBooleanOption(
+        semicolon, other.semicolon, "Semicolon()");
+
+    // Whenever we combine two TextOptions, we reset the headers and
+    // datasetInfo.
+    headers.clear();
+    datasetInfo = DatasetInfo();
+  }
+
   // Print warnings for any members that cannot be represented by a
   // DataOptionsBase<void>.
   void WarnBaseConversion(const char* dataDescription) const
@@ -133,7 +181,9 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     semicolon.reset();
     missingToNan.reset();
     categorical.reset();
-    missingPolicy.reset();
+
+    headers.clear();
+    datasetInfo = DatasetInfo();
   }
 
   // Get if the dataset has headers or not.
@@ -153,7 +203,7 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     return this->AccessMember(semicolon, defaultSemicolon);
   }
   // Modify the separator type in the matrix.
-  bool& SemiColon()
+  bool& Semicolon()
   {
     return this->ModifyMember(semicolon, defaultSemicolon);
   }
@@ -182,63 +232,51 @@ class TextOptions : public MatrixOptionsBase<TextOptions>
     return this->ModifyMember(categorical, defaultCategorical);
   }
 
-  // Get whether the data should be interpreted as categorical when columns are
-  // not numeric.
-  bool MissingPolicy() const
-  {
-    return this->AccessMember(missingPolicy, defaultMissingPolicy);
-  }
-  // Modify whether the data should be interpreted as missingPolicy when columns
-  // are not numeric.
-  bool& MissingPolicy()
-  {
-    return this->ModifyMember(missingPolicy, defaultMissingPolicy);
-  }
-
   // Get the headers.
   const arma::field<std::string>& Headers() const { return headers; }
   // Modify the headers.
   arma::field<std::string>& Headers() { return headers; }
 
   // Get the DatasetInfo for categorical data.
-  const data::DatasetInfo& DatasetInfo() const { return datasetInfo; }
-  // Modify the DatasetInfo.
-  data::DatasetInfo& DatasetInfo() { return datasetInfo; }
-
-  // Get the DatasetInfo for categorical data.
-  const data::DatasetMapper<data::MissingPolicy>& DatasetMissingPolicy() const
+  const DatasetMapper<IncrementPolicy>& DatasetInfo() const
   {
-    return datasetMissingPolicy;
+    return datasetInfo;
   }
 
-  // Modify the DatasetMissingPolicy.
-  data::DatasetMapper<data::MissingPolicy>& DatasetMissingPolicy()
-  {
-    return datasetMissingPolicy;
-  }
+  // Modify the DatasetInfo.
+  DatasetMapper<IncrementPolicy>& DatasetInfo() { return datasetInfo; }
 
  private:
   std::optional<bool> hasHeaders;
   std::optional<bool> semicolon;
   std::optional<bool> missingToNan;
   std::optional<bool> categorical;
-  std::optional<bool> missingPolicy;
 
   // These are not optional, but if either is specified, then it should be taken
   // to mean that `hasHeaders` or `categorical` has been specified as true.
   arma::field<std::string> headers;
-  data::DatasetInfo datasetInfo;
-  // Temporary internal member until MissingPolicy is refactored out.
-  data::DatasetMapper<data::MissingPolicy> datasetMissingPolicy;
+  DatasetMapper<IncrementPolicy> datasetInfo;
 
   constexpr static const bool defaultHasHeaders = false;
   constexpr static const bool defaultSemicolon = false;
   constexpr static const bool defaultMissingToNan = false;
   constexpr static const bool defaultCategorical = false;
-  constexpr static const bool defaultMissingPolicy = false;
 };
 
-} // namespace data
+// Boolean options
+static const TextOptions HasHeaders   = TextOptions(true);
+static const TextOptions Semicolon    = TextOptions(std::nullopt, true);
+static const TextOptions MissingToNan = TextOptions(std::nullopt, std::nullopt,
+      true);
+static const TextOptions Categorical  = TextOptions(std::nullopt,
+      std::nullopt, std::nullopt, true);
+
+template<>
+struct IsDataOptions<TextOptions>
+{
+  constexpr static bool value = true;
+};
+
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/tokenizers/char_extract.hpp

@@ -17,7 +17,6 @@
 #include <mlpack/prereqs.hpp>
 
 namespace mlpack {
-namespace data {
 
 /**
  * The class is used to split a string into characters.
@@ -59,7 +58,6 @@ class CharExtract
   }
 };
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/data/tokenizers/split_by_any_of.hpp

@@ -18,7 +18,6 @@
 #include <array>
 
 namespace mlpack {
-namespace data {
 
 /**
  * The SplitByAnyOf class tokenizes a string using a set of delimiters.
@@ -112,7 +111,6 @@ class SplitByAnyOf
   MaskType mask;
 };
 
-} // namespace data
 } // namespace mlpack
 
 #endif

mlpack/include/mlpack/core/distributions/gamma_distribution_impl.hpp

@@ -59,12 +59,12 @@ inline void GammaDistribution<MatType>::Train(const MatType& rdata,
                                               const ElemType tol)
 {
   // If fittingSet is empty, nothing to do.
-  if (arma::size(rdata) == arma::size(arma::mat()))
+  if (rdata.is_empty())
     return;
 
   // Calculate log(mean(x)) and mean(log(x)) of each dataset row.
-  const VecType meanLogxVec = arma::mean(log(rdata), 1);
-  const VecType meanxVec = arma::mean(rdata, 1);
+  const VecType meanLogxVec = mean(log(rdata), 1);
+  const VecType meanxVec = mean(rdata, 1);
   const VecType logMeanxVec = log(meanxVec);
 
   // Call the statistics-only GammaDistribution::Train() function to fit the
@@ -79,7 +79,7 @@ inline void GammaDistribution<MatType>::Train(const MatType& rdata,
                                               const ElemType tol)
 {
   // If fittingSet is empty, nothing to do.
-  if (arma::size(rdata) == arma::size(arma::mat()))
+  if (rdata.is_empty())
     return;
 
   VecType meanLogxVec(rdata.n_rows);