RecFunctions

class numpy::RecFunctions

numpy C++ class.

Example

#include <numpy/np_ndarray.h>
using namespace numpy;

// Use RecFunctions
RecFunctions obj;
// ... operations ...

Construction

Signature	Return Type	Location	Example
StructuredValue createComparisonValue(const StructuredRecord &record)	StructuredValue	NP_RECFUNCTIONS.H:535

Indexing / Selection

Signature	Return Type	Location	Example
std::string get_fieldstructure(const RecordArray &array)	std::string	NP_RECFUNCTIONS.H:412	View

Joining / Splitting

Signature	Return Type	Location	Example
RecordArray append_fields(const RecordArray &base, const std::vector<std::string>&names, const std::vector<std::pair<DType,std::shared_ptr<void>>>&data, const std::vector<StructuredValue>&fill_values = {}, boolusemask = false, boolasrecarray = true)	RecordArray	NP_RECFUNCTIONS.H:13	View
RecordArray stack_arrays(const std::vector<RecordArray>&arrays, booldefaults = false, boolusemask = false, boolasrecarray = true, boolautoconvert = false)	RecordArray	NP_RECFUNCTIONS.H:217	View

Other Methods

Signature	Return Type	Location	Example
size_t calculateLinearIndex(const std::vector<size_t>&indices, const std::vector<size_t>&shape)	size_t	NP_RECFUNCTIONS.H:498
RecordArray drop_fields(const RecordArray &base, const std::vector<std::string>&drop_names, boolusemask = false, boolasrecarray = true)	RecordArray	NP_RECFUNCTIONS.H:62	View
RecordArray join_by(const std::string &key, const RecordArray &r1, const RecordArray &r2, const std::string &jointype = "inner", const std::string &r1postfix = "1", const std::string &r2postfix = "2", booldefaults = false, boolusemask = false, boolasrecarray = true)	RecordArray	NP_RECFUNCTIONS.H:270	View
RecordArray merge_arrays(const std::vector<RecordArray>&arrays, boolflatten = false, boolusemask = false, boolasrecarray = true)	RecordArray	NP_RECFUNCTIONS.H:147	View
RecordArray rename_fields(const RecordArray &base, const std::unordered_map<std::string,std::string>&namemap)	RecordArray	NP_RECFUNCTIONS.H:105	View
RecordArray repack_fields(const RecordArray &array, boolalign = false)	RecordArray	NP_RECFUNCTIONS.H:416	View
void setFieldFromData(RecordArray &array, const std::vector<size_t>&indices, const std::string &field_name, const std::pair<DType,std::shared_ptr<void>>&data, size_toffset)	void	NP_RECFUNCTIONS.H:508

Code Examples

The following examples are extracted from the test suite.

get_fieldstructure (np_test_1_all.cpp:21735)

    // std::cout << "Base array:" << std::endl;
    //base.printArray();

    std::vector<std::string> names = get_names(base);
    // std::cout << "Field names: ";
    for (const auto& name : names) {
      // std::cout << name << " ";
    }
    // std::cout << std::endl;

    // std::cout << "Field structure: " << get_fieldstructure(base) << std::endl;

    auto dropped = drop_fields(base, { "score" });
    // std::cout << "After dropping 'score' field:";
    //dropped.printArray();

    std::unordered_map<std::string, std::string> rename_map = { {"name", "student_name"} };
    auto renamed = rename_fields(base, rename_map);
    // std::cout << "After renaming 'name' to 'student_name':" << std::endl;
    //renamed.printArray();

append_fields (np_test_1_all.cpp:25911)

      auto score_ptr = make_shared_data(score_data);
      auto rank_ptr = make_shared_data(rank_data);

      // Append fields
      std::vector<std::string> new_field_names = { "score", "rank" };
      std::vector<std::pair<numpy::DType, std::shared_ptr<void>>> new_field_data = {
          {numpy::DType::FLOAT64, score_ptr},
          {numpy::DType::INT32, rank_ptr}
      };

      auto result = numpy::RecFunctions::append_fields(base_arr, new_field_names, new_field_data);

      // Verify result has 4 fields
      auto field_names = result.getFieldNames();
      bool passed = (field_names.size() == 4);

      if (passed) {
        // Verify original fields preserved
        passed = (result.getFieldValue<numpy::int32>({ 0 }, "id") == 1);
        passed = passed && (result.getFieldValue<numpy::int32>({ 1 }, "id") == 2);
        passed = passed && (result.getFieldValue<numpy::int32>({ 2 }, "name_length") == 6);

stack_arrays (np_test_1_all.cpp:21784)

    // std::cout << "Array 1:" << std::endl;
    //arr1.printArray();

    // std::cout << "Array 2:" << std::endl;
    //arr2.printArray();

    auto merged = merge_arrays({ arr1, arr2 });
    // std::cout << "Merged arrays:";
    //merged.printArray();

    auto stacked = stack_arrays({ arr1 });
    // std::cout << "Stacked arrays (single array):";
    //stacked.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

  void testRecFunctionsJoin() {
    std::cout << "========= testRecFunctionsJoin =======================";

    auto dtype = std::make_shared<StructuredDType>(std::vector<std::pair<std::string, DType>>{

drop_fields (np_test_1_all.cpp:21737)

    std::vector<std::string> names = get_names(base);
    // std::cout << "Field names: ";
    for (const auto& name : names) {
      // std::cout << name << " ";
    }
    // std::cout << std::endl;

    // std::cout << "Field structure: " << get_fieldstructure(base) << std::endl;

    auto dropped = drop_fields(base, { "score" });
    // std::cout << "After dropping 'score' field:";
    //dropped.printArray();

    std::unordered_map<std::string, std::string> rename_map = { {"name", "student_name"} };
    auto renamed = rename_fields(base, rename_map);
    // std::cout << "After renaming 'name' to 'student_name':" << std::endl;
    //renamed.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

join_by (np_test_1_all.cpp:21826)

    right.setFieldValue({ 1 }, "info", std::string("B"));
    right.setFieldValue({ 2 }, "key", static_cast<int32>(4));
    right.setFieldValue({ 2 }, "info", std::string("D"));

    // std::cout << "Left array:" << std::endl;
    //left.printArray();

    // std::cout << "Right array:" << std::endl;
    //right.printArray();

    auto inner_join = join_by("key", left, right, "inner");
    // std::cout << "Inner join on 'key':" << std::endl;
    //inner_join.printArray();

    auto outer_join = join_by("key", left, right, "outer");
    // std::cout << "Outer join on 'key':" << std::endl;
    //outer_join.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

merge_arrays (np_test_1_all.cpp:21780)

    arr2.setFieldValue({ 0 }, "name", std::string("Alice"));
    arr2.setFieldValue({ 1 }, "id", static_cast<int32>(4));
    arr2.setFieldValue({ 1 }, "name", std::string("Bob"));

    // std::cout << "Array 1:" << std::endl;
    //arr1.printArray();

    // std::cout << "Array 2:" << std::endl;
    //arr2.printArray();

    auto merged = merge_arrays({ arr1, arr2 });
    // std::cout << "Merged arrays:";
    //merged.printArray();

    auto stacked = stack_arrays({ arr1 });
    // std::cout << "Stacked arrays (single array):";
    //stacked.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

rename_fields (np_test_1_all.cpp:21742)

    }
    // std::cout << std::endl;

    // std::cout << "Field structure: " << get_fieldstructure(base) << std::endl;

    auto dropped = drop_fields(base, { "score" });
    // std::cout << "After dropping 'score' field:";
    //dropped.printArray();

    std::unordered_map<std::string, std::string> rename_map = { {"name", "student_name"} };
    auto renamed = rename_fields(base, rename_map);
    // std::cout << "After renaming 'name' to 'student_name':" << std::endl;
    //renamed.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

  void testRecFunctionsMergeStack() {
    std::cout << "========= testRecFunctionsMergeStack =======================";

    auto dtype1 = std::make_shared<StructuredDType>(std::vector<std::pair<std::string, DType>>{

repack_fields (np_test_1_all.cpp:21867)

    // std::cout << "Array with duplicates:" << std::endl;
    //arr.printArray();

    auto duplicates = find_duplicates(arr, "id");
    // std::cout << "Duplicate indices based on 'id' field: ";
    for (size_t idx : duplicates) {
      // std::cout << idx << " ";
    }
    // std::cout << std::endl;

    auto repacked = repack_fields(arr, false);
    // std::cout << "Repacked array:" << std::endl;
    //repacked.printArray();

    std::cout << " -> tests passed" << std::endl;
  }

  void testStringParsing() {
    std::cout << "========= testStringParsing =======================";

    auto dtype = std::make_shared<StructuredDType>(std::vector<std::pair<std::string, DType>>{