nmatrix 0.0.8 → 0.0.9

data/ext/nmatrix/storage/yale/yale.cpp

@@ -0,0 +1,1785 @@
+/////////////////////////////////////////////////////////////////////
+// = NMatrix
+//
+// A linear algebra library for scientific computation in Ruby.
+// NMatrix is part of SciRuby.
+//
+// NMatrix was originally inspired by and derived from NArray, by
+// Masahiro Tanaka: http://narray.rubyforge.org
+//
+// == Copyright Information
+//
+// SciRuby is Copyright (c) 2010 - 2013, Ruby Science Foundation
+// NMatrix is Copyright (c) 2013, Ruby Science Foundation
+//
+// Please see LICENSE.txt for additional copyright notices.
+//
+// == Contributing
+//
+// By contributing source code to SciRuby, you agree to be bound by
+// our Contributor Agreement:
+//
+// * https://github.com/SciRuby/sciruby/wiki/Contributor-Agreement
+//
+// == yale.c
+//
+// "new yale" storage format for 2D matrices (like yale, but with
+// the diagonal pulled out for O(1) access).
+//
+// Specifications:
+// * dtype and index dtype must necessarily differ
+//      * index dtype is defined by whatever unsigned type can store
+//        max(rows,cols)
+//      * that means vector ija stores only index dtype, but a stores
+//        dtype
+// * vectors must be able to grow as necessary
+//      * maximum size is rows*cols+1
+
+/*
+ * Standard Includes
+ */
+
+#include <ruby.h>
+#include <algorithm>  // std::min
+#include <cstdio>     // std::fprintf
+#include <iostream>
+#include <array>
+#include <typeinfo>
+#include <tuple>
+#include <queue>
+
+/*
+ * Project Includes
+ */
+
+// #include "types.h"
+#include "../../data/data.h"
+#include "../../math/math.h"
+
+#include "../common.h"
+
+#include "../../nmatrix.h"
+#include "../../data/meta.h"
+
+#include "iterators/base.h"
+#include "iterators/stored_diagonal.h"
+#include "iterators/row_stored_nd.h"
+#include "iterators/row_stored.h"
+#include "iterators/row.h"
+#include "iterators/iterator.h"
+#include "class.h"
+#include "yale.h"
+#include "../../ruby_constants.h"
+
+/*
+ * Macros
+ */
+
+#ifndef NM_MAX
+#define NM_MAX(a,b) (((a)>(b))?(a):(b))
+#define NM_MIN(a,b) (((a)<(b))?(a):(b))
+#endif
+
+/*
+ * Forward Declarations
+ */
+
+extern "C" {
+  static YALE_STORAGE*	alloc(nm::dtype_t dtype, size_t* shape, size_t dim);
+
+  static size_t yale_count_slice_copy_ndnz(const YALE_STORAGE* s, size_t*, size_t*);
+
+  static void* default_value_ptr(const YALE_STORAGE* s);
+  static VALUE default_value(const YALE_STORAGE* s);
+  static VALUE obj_at(YALE_STORAGE* s, size_t k);
+
+  /* Ruby-accessible functions */
+  static VALUE nm_size(VALUE self);
+  static VALUE nm_a(int argc, VALUE* argv, VALUE self);
+  static VALUE nm_d(int argc, VALUE* argv, VALUE self);
+  static VALUE nm_lu(VALUE self);
+  static VALUE nm_ia(VALUE self);
+  static VALUE nm_ja(VALUE self);
+  static VALUE nm_ija(int argc, VALUE* argv, VALUE self);
+
+  static VALUE nm_nd_row(int argc, VALUE* argv, VALUE self);
+
+  static inline size_t src_ndnz(const YALE_STORAGE* s) {
+    return reinterpret_cast<YALE_STORAGE*>(s->src)->ndnz;
+  }
+
+} // end extern "C" block
+
+namespace nm { namespace yale_storage {
+
+template <typename LD, typename RD>
+static VALUE map_merged_stored(VALUE left, VALUE right, VALUE init);
+
+template <typename DType>
+static bool						ndrow_is_empty(const YALE_STORAGE* s, IType ija, const IType ija_next);
+
+template <typename LDType, typename RDType>
+static bool						ndrow_eqeq_ndrow(const YALE_STORAGE* l, const YALE_STORAGE* r, IType l_ija, const IType l_ija_next, IType r_ija, const IType r_ija_next);
+
+template <typename LDType, typename RDType>
+static bool           eqeq(const YALE_STORAGE* left, const YALE_STORAGE* right);
+
+template <typename LDType, typename RDType>
+static bool eqeq_different_defaults(const YALE_STORAGE* s, const LDType& s_init, const YALE_STORAGE* t, const RDType& t_init);
+
+static void						increment_ia_after(YALE_STORAGE* s, IType ija_size, IType i, long n);
+
+static IType				  insert_search(YALE_STORAGE* s, IType left, IType right, IType key, bool& found);
+
+template <typename DType>
+static char           vector_insert(YALE_STORAGE* s, size_t pos, size_t* j, void* val_, size_t n, bool struct_only);
+
+template <typename DType>
+static char           vector_insert_resize(YALE_STORAGE* s, size_t current_size, size_t pos, size_t* j, size_t n, bool struct_only);
+
+template <typename DType>
+static std::tuple<long,bool,std::queue<std::tuple<IType,IType,int> > > count_slice_set_ndnz_change(YALE_STORAGE* s, size_t* coords, size_t* lengths, DType* v, size_t v_size);
+
+static inline IType* IJA(const YALE_STORAGE* s) {
+  return reinterpret_cast<YALE_STORAGE*>(s->src)->ija;
+}
+
+static inline IType IJA_SET(const YALE_STORAGE* s, size_t loc, IType val) {
+  return IJA(s)[loc] = val;
+}
+
+template <typename DType>
+static inline DType* A(const YALE_STORAGE* s) {
+  return reinterpret_cast<DType*>(reinterpret_cast<YALE_STORAGE*>(s->src)->a);
+}
+
+template <typename DType>
+static inline DType A_SET(const YALE_STORAGE* s, size_t loc, DType val) {
+  return A<DType>(s)[loc] = val;
+}
+
+
+/*
+ * Functions
+ */
+
+/*
+ * Copy a vector from one DType to another.
+ */
+template <typename LType, typename RType>
+static inline void copy_recast_vector(const void* in_, void* out_, size_t length) {
+  const RType* in = reinterpret_cast<const RType*>(in_);
+  LType* out      = reinterpret_cast<LType*>(out_);
+  for (size_t i = 0; i < length; ++i) {
+    out[i] = in[i];
+  }
+  out;
+}
+
+
+
+/*
+ * Create Yale storage from IA, JA, and A vectors given in Old Yale format (probably from a file, since NMatrix only uses
+ * new Yale for its storage).
+ *
+ * This function is needed for Matlab .MAT v5 IO.
+ */
+template <typename LDType, typename RDType>
+YALE_STORAGE* create_from_old_yale(dtype_t dtype, size_t* shape, char* r_ia, char* r_ja, char* r_a) {
+  IType*  ir = reinterpret_cast<IType*>(r_ia);
+  IType*  jr = reinterpret_cast<IType*>(r_ja);
+  RDType* ar = reinterpret_cast<RDType*>(r_a);
+
+  // Read through ia and ja and figure out the ndnz (non-diagonal non-zeros) count.
+  size_t ndnz = 0, i, p, p_next;
+
+  for (i = 0; i < shape[0]; ++i) { // Walk down rows
+    for (p = ir[i], p_next = ir[i+1]; p < p_next; ++p) { // Now walk through columns
+
+      if (i != jr[p]) ++ndnz; // entry is non-diagonal and probably nonzero
+
+    }
+  }
+
+  // Having walked through the matrix, we now go about allocating the space for it.
+  YALE_STORAGE* s = alloc(dtype, shape, 2);
+
+  s->capacity = shape[0] + ndnz + 1;
+  s->ndnz     = ndnz;
+
+  // Setup IJA and A arrays
+  s->ija = ALLOC_N( IType, s->capacity );
+  s->a   = ALLOC_N( LDType, s->capacity );
+  IType* ijl    = reinterpret_cast<IType*>(s->ija);
+  LDType* al    = reinterpret_cast<LDType*>(s->a);
+
+  // set the diagonal to zero -- this prevents uninitialized values from popping up.
+  for (size_t index = 0; index < shape[0]; ++index) {
+    al[index] = 0;
+  }
+
+  // Figure out where to start writing JA in IJA:
+  size_t pp = s->shape[0]+1;
+
+  // Find beginning of first row
+  p = ir[0];
+
+  // Now fill the arrays
+  for (i = 0; i < s->shape[0]; ++i) {
+
+    // Set the beginning of the row (of output)
+    ijl[i] = pp;
+
+    // Now walk through columns, starting at end of row (of input)
+    for (size_t p_next = ir[i+1]; p < p_next; ++p, ++pp) {
+
+      if (i == jr[p]) { // diagonal
+
+        al[i] = ar[p];
+        --pp;
+
+      } else {          // nondiagonal
+
+        ijl[pp] = jr[p];
+        al[pp]  = ar[p];
+
+      }
+    }
+  }
+
+  ijl[i] = pp; // Set the end of the last row
+
+  // Set the zero position for our output matrix
+  al[i] = 0;
+
+  return s;
+}
+
+
+/*
+ * Empty the matrix by initializing the IJA vector and setting the diagonal to 0.
+ *
+ * Called when most YALE_STORAGE objects are created.
+ *
+ * Can't go inside of class YaleStorage because YaleStorage creation requires that
+ * IJA already be initialized.
+ */
+template <typename DType>
+void init(YALE_STORAGE* s, void* init_val) {
+  IType IA_INIT = s->shape[0] + 1;
+
+  IType* ija = reinterpret_cast<IType*>(s->ija);
+  // clear out IJA vector
+  for (IType i = 0; i < IA_INIT; ++i) {
+    ija[i] = IA_INIT; // set initial values for IJA
+  }
+
+  clear_diagonal_and_zero<DType>(s, init_val);
+}
+
+
+template <typename LDType, typename RDType>
+static YALE_STORAGE* slice_copy(YALE_STORAGE* s) {
+  YaleStorage<RDType> y(s);
+  return y.template alloc_copy<LDType, false>();
+}
+
+
+/*
+ * Template version of copy transposed. This could also, in theory, allow a map -- but transpose.h
+ * would need to be updated.
+ *
+ * TODO: Update for slicing? Update for different dtype in and out? We can cast rather easily without
+ * too much modification.
+ */
+template <typename D>
+YALE_STORAGE* copy_transposed(YALE_STORAGE* rhs) {
+  YaleStorage<D> y(rhs);
+  return y.template alloc_copy_transposed<D, false>();
+}
+
+
+///////////////
+// Accessors //
+///////////////
+
+
+/*
+ * Determine the number of non-diagonal non-zeros in a not-yet-created copy of a slice or matrix.
+ */
+template <typename DType>
+static size_t count_slice_copy_ndnz(const YALE_STORAGE* s, size_t* offset, size_t* shape) {
+  IType* ija = s->ija;
+  DType* a   = reinterpret_cast<DType*>(s->a);
+
+  DType ZERO(*reinterpret_cast<DType*>(default_value_ptr(s)));
+
+  // Calc ndnz for the destination
+  size_t ndnz  = 0;
+  size_t i, j; // indexes of destination matrix
+  size_t k, l; // indexes of source matrix
+  for (i = 0; i < shape[0]; i++) {
+    k = i + offset[0];
+    for (j = 0; j < shape[1]; j++) {
+      l = j + offset[1];
+
+      if (j == i)  continue;
+
+      if (k == l) { // for diagonal element of source
+        if (a[k] != ZERO) ++ndnz;
+      } else { // for non-diagonal element
+        for (size_t c = ija[k]; c < ija[k+1]; c++) {
+          if (ija[c] == l) {
+            ++ndnz;
+            break;
+          }
+        }
+      }
+    }
+  }
+
+  return ndnz;
+}
+
+
+
+/*
+ * Get a single element of a yale storage object
+ */
+template <typename DType>
+static void* get_single(YALE_STORAGE* storage, SLICE* slice) {
+  YaleStorage<DType> y(storage);
+  return reinterpret_cast<void*>(y.get_single_p(slice));
+}
+
+
+/*
+ * Returns a reference-slice of a matrix.
+ */
+template <typename DType>
+YALE_STORAGE* ref(YALE_STORAGE* s, SLICE* slice) {
+  return YaleStorage<DType>(s).alloc_ref(slice);
+}
+
+
+/*
+ * Attempt to set a cell or cells in a Yale matrix.
+ */
+template <typename DType>
+void set(VALUE left, SLICE* slice, VALUE right) {
+  YALE_STORAGE* storage = NM_STORAGE_YALE(left);
+  YaleStorage<DType> y(storage);
+  y.insert(slice, right);
+}
+
+///////////
+// Tests //
+///////////
+
+/*
+ * Yale eql? -- for whole-matrix comparison returning a single value.
+ */
+template <typename LDType, typename RDType>
+static bool eqeq(const YALE_STORAGE* left, const YALE_STORAGE* right) {
+  return YaleStorage<LDType>(left) == YaleStorage<RDType>(right);
+}
+
+
+//////////
+// Math //
+//////////
+
+#define YALE_IA(s) (reinterpret_cast<IType*>(s->ija))
+#define YALE_IJ(s) (reinterpret_cast<IType*>(s->ija) + s->shape[0] + 1)
+#define YALE_COUNT(yale) (yale->ndnz + yale->shape[0])
+
+/////////////
+// Utility //
+/////////////
+
+
+/*
+ * Binary search for finding the beginning of a slice. Returns the position of the first element which is larger than
+ * bound.
+ */
+IType binary_search_left_boundary(const YALE_STORAGE* s, IType left, IType right, IType bound) {
+  if (left > right) return -1;
+
+  IType* ija  = IJA(s);
+
+  if (ija[left] >= bound) return left; // shortcut
+
+  IType mid   = (left + right) / 2;
+  IType mid_j = ija[mid];
+
+  if (mid_j == bound)
+    return mid;
+  else if (mid_j > bound) { // eligible! don't exclude it.
+    return binary_search_left_boundary(s, left, mid, bound);
+  } else // (mid_j < bound)
+    return binary_search_left_boundary(s, mid + 1, right, bound);
+}
+
+
+/*
+ * Binary search for returning stored values. Returns a non-negative position, or -1 for not found.
+ */
+int binary_search(YALE_STORAGE* s, IType left, IType right, IType key) {
+  if (s->src != s) throw; // need to fix this quickly
+
+  if (left > right) return -1;
+
+  IType* ija = s->ija;
+
+  IType mid = (left + right)/2;
+  IType mid_j = ija[mid];
+
+  if (mid_j == key)
+  	return mid;
+
+  else if (mid_j > key)
+  	return binary_search(s, left, mid - 1, key);
+
+  else
+  	return binary_search(s, mid + 1, right, key);
+}
+
+
+/*
+ * Resize yale storage vectors A and IJA, copying values.
+ */
+static void vector_grow(YALE_STORAGE* s) {
+  if (s != s->src) {
+    throw; // need to correct this quickly.
+  }
+
+  size_t new_capacity = s->capacity * GROWTH_CONSTANT;
+  size_t max_capacity = YaleStorage<uint8_t>::max_size(s->shape);
+
+  if (new_capacity > max_capacity) new_capacity = max_capacity;
+
+  IType* new_ija      = ALLOC_N(IType, new_capacity);
+  void* new_a         = ALLOC_N(char, DTYPE_SIZES[s->dtype] * new_capacity);
+
+  IType* old_ija      = s->ija;
+  void* old_a         = s->a;
+
+  memcpy(new_ija, old_ija, s->capacity * sizeof(IType));
+  memcpy(new_a,   old_a,   s->capacity * DTYPE_SIZES[s->dtype]);
+
+  s->capacity         = new_capacity;
+
+  xfree(old_ija);
+  xfree(old_a);
+
+  s->ija         = new_ija;
+  s->a           = new_a;
+}
+
+
+/*
+ * Resize yale storage vectors A and IJA in preparation for an insertion.
+ */
+template <typename DType>
+static char vector_insert_resize(YALE_STORAGE* s, size_t current_size, size_t pos, size_t* j, size_t n, bool struct_only) {
+  if (s != s->src) throw;
+
+  // Determine the new capacity for the IJA and A vectors.
+  size_t new_capacity = s->capacity * GROWTH_CONSTANT;
+  size_t max_capacity = YaleStorage<DType>::max_size(s->shape);
+
+  if (new_capacity > max_capacity) {
+    new_capacity = max_capacity;
+
+    if (current_size + n > max_capacity) rb_raise(rb_eNoMemError, "insertion size exceeded maximum yale matrix size");
+  }
+
+  if (new_capacity < current_size + n)
+  	new_capacity = current_size + n;
+
+  // Allocate the new vectors.
+  IType* new_ija     = ALLOC_N( IType, new_capacity );
+  NM_CHECK_ALLOC(new_ija);
+
+  DType* new_a       = ALLOC_N( DType, new_capacity );
+  NM_CHECK_ALLOC(new_a);
+
+  IType* old_ija     = reinterpret_cast<IType*>(s->ija);
+  DType* old_a       = reinterpret_cast<DType*>(s->a);
+
+  // Copy all values prior to the insertion site to the new IJA and new A
+  if (struct_only) {
+    for (size_t i = 0; i < pos; ++i) {
+      new_ija[i] = old_ija[i];
+    }
+  } else {
+    for (size_t i = 0; i < pos; ++i) {
+      new_ija[i] = old_ija[i];
+      new_a[i]   = old_a[i];
+    }
+  }
+
+
+  // Copy all values subsequent to the insertion site to the new IJA and new A, leaving room (size n) for insertion.
+  if (struct_only) {
+    for (size_t i = pos; i < current_size; ++i) {
+      new_ija[i+n] = old_ija[i];
+    }
+  } else {
+    for (size_t i = pos; i < current_size; ++i) {
+      new_ija[i+n] = old_ija[i];
+      new_a[i+n] = old_a[i];
+    }
+  }
+
+  s->capacity = new_capacity;
+
+  xfree(s->ija);
+  xfree(s->a);
+
+  s->ija = new_ija;
+  s->a   = reinterpret_cast<void*>(new_a);
+
+  return 'i';
+}
+
+/*
+ * Insert a value or contiguous values in the ija and a vectors (after ja and
+ * diag). Does not free anything; you are responsible!
+ *
+ * TODO: Improve this so it can handle non-contiguous element insertions
+ *	efficiently. For now, we can just sort the elements in the row in
+ *	question.)
+ */
+template <typename DType>
+static char vector_insert(YALE_STORAGE* s, size_t pos, size_t* j, void* val_, size_t n, bool struct_only) {
+
+  if (pos < s->shape[0]) {
+    rb_raise(rb_eArgError, "vector insert pos (%lu) is before beginning of ja (%lu); this should not happen", pos, s->shape[0]);
+  }
+
+  DType* val = reinterpret_cast<DType*>(val_);
+
+  size_t size = s->ija[s->shape[0]];
+
+  IType* ija = s->ija;
+  DType* a   = reinterpret_cast<DType*>(s->a);
+
+  if (size + n > s->capacity) {
+  	vector_insert_resize<DType>(s, size, pos, j, n, struct_only);
+
+    // Need to get the new locations for ija and a.
+  	ija = s->ija;
+    a   = reinterpret_cast<DType*>(s->a);
+
+  } else {
+    /*
+     * No resize required:
+     * easy (but somewhat slow), just copy elements to the tail, starting at
+     * the end, one element at a time.
+     *
+     * TODO: This can be made slightly more efficient, but only after the tests
+     *	are written.
+     */
+
+    if (struct_only) {
+      for (size_t i = 0; i < size - pos; ++i) {
+        ija[size+n-1-i] = ija[size-1-i];
+      }
+    } else {
+      for (size_t i = 0; i < size - pos; ++i) {
+        ija[size+n-1-i] = ija[size-1-i];
+        a[size+n-1-i]   = a[size-1-i];
+      }
+    }
+  }
+
+  // Now insert the new values.
+  if (struct_only) {
+    for (size_t i = 0; i < n; ++i) {
+      ija[pos+i]  = j[i];
+    }
+  } else {
+    for (size_t i = 0; i < n; ++i) {
+      ija[pos+i]  = j[i];
+      a[pos+i]    = val[i];
+    }
+  }
+
+  return 'i';
+}
+
+/*
+ * If we add n items to row i, we need to increment ija[i+1] and onward.
+ */
+static void increment_ia_after(YALE_STORAGE* s, IType ija_size, IType i, long n) {
+  IType* ija = s->ija;
+
+  ++i;
+  for (; i <= ija_size; ++i) {
+    ija[i] += n;
+  }
+}
+
+/*
+ * Binary search for returning insertion points.
+ */
+static IType insert_search(YALE_STORAGE* s, IType left, IType right, IType key, bool& found) {
+
+  if (left > right) {
+    found = false;
+    return left;
+  }
+
+  IType* ija = s->ija;
+  IType mid = (left + right)/2;
+  IType mid_j = ija[mid];
+
+  if (mid_j == key) {
+    found = true;
+    return mid;
+
+  } else if (mid_j > key) {
+  	return insert_search(s, left, mid-1, key, found);
+
+  } else {
+  	return insert_search(s, mid+1, right, key, found);
+  }
+}
+
+/////////////////////////
+// Copying and Casting //
+/////////////////////////
+
+/*
+ * Templated copy constructor for changing dtypes.
+ */
+template <typename L, typename R>
+YALE_STORAGE* cast_copy(const YALE_STORAGE* rhs) {
+  YaleStorage<R> y(rhs);
+  return y.template alloc_copy<L>();
+}
+
+/*
+ * Template access for getting the size of Yale storage.
+ */
+size_t get_size(const YALE_STORAGE* storage) {
+  return storage->ija[ storage->shape[0] ];
+}
+
+
+template <typename DType>
+static STORAGE* matrix_multiply(const STORAGE_PAIR& casted_storage, size_t* resulting_shape, bool vector) {
+  YALE_STORAGE *left  = (YALE_STORAGE*)(casted_storage.left),
+               *right = (YALE_STORAGE*)(casted_storage.right);
+
+  // We can safely get dtype from the casted matrices; post-condition of binary_storage_cast_alloc is that dtype is the
+  // same for left and right.
+  // int8_t dtype = left->dtype;
+
+  IType* ijl = left->ija;
+  IType* ijr = right->ija;
+
+  // First, count the ndnz of the result.
+  // TODO: This basically requires running symbmm twice to get the exact ndnz size. That's frustrating. Are there simple
+  // cases where we can avoid running it?
+  size_t result_ndnz = nm::math::symbmm(resulting_shape[0], left->shape[1], resulting_shape[1], ijl, ijl, true, ijr, ijr, true, NULL, true);
+
+  // Create result storage.
+  YALE_STORAGE* result = nm_yale_storage_create(left->dtype, resulting_shape, 2, result_ndnz);
+  init<DType>(result, NULL);
+  IType* ija = result->ija;
+
+  // Symbolic multiplication step (build the structure)
+  nm::math::symbmm(resulting_shape[0], left->shape[1], resulting_shape[1], ijl, ijl, true, ijr, ijr, true, ija, true);
+
+  // Numeric multiplication step (fill in the elements)
+
+  nm::math::numbmm<DType>(result->shape[0], left->shape[1], result->shape[1],
+                                ijl, ijl, reinterpret_cast<DType*>(left->a), true,
+                                ijr, ijr, reinterpret_cast<DType*>(right->a), true,
+                                ija, ija, reinterpret_cast<DType*>(result->a), true);
+
+
+  // Sort the columns
+  nm::math::smmp_sort_columns<DType>(result->shape[0], ija, ija, reinterpret_cast<DType*>(result->a));
+
+  return reinterpret_cast<STORAGE*>(result);
+}
+
+
+/*
+ * Get the sum of offsets from the original matrix (for sliced iteration).
+ */
+static std::array<size_t,2> get_offsets(YALE_STORAGE* x) {
+  std::array<size_t, 2> offsets{ {0,0} };
+  while (x != x->src) {
+    offsets[0] += x->offset[0];
+    offsets[1] += x->offset[1];
+    x = reinterpret_cast<YALE_STORAGE*>(x->src);
+  }
+  return offsets;
+}
+
+
+class RowIterator {
+protected:
+  YALE_STORAGE* s;
+  IType* ija;
+  void*  a;
+  IType i, k, k_end;
+  size_t j_offset, j_shape;
+  bool diag, End;
+  VALUE init;
+public:
+  RowIterator(YALE_STORAGE* s_, IType* ija_, IType i_, size_t j_shape_, size_t j_offset_ = 0)
+    : s(s_),
+      ija(ija_),
+      a(reinterpret_cast<YALE_STORAGE*>(s->src)->a),
+      i(i_),
+      k(ija[i]),
+      k_end(ija[i+1]),
+      j_offset(j_offset_),
+      j_shape(j_shape_),
+      diag(row_has_no_nd() || diag_is_first()),
+      End(false),
+      init(default_value(s))
+    { }
+
+  RowIterator(YALE_STORAGE* s_, IType i_, size_t j_shape_, size_t j_offset_ = 0)
+    : s(s_),
+      ija(IJA(s)),
+      a(reinterpret_cast<YALE_STORAGE*>(s->src)->a),
+      i(i_),
+      k(ija[i]),
+      k_end(ija[i+1]),
+      j_offset(j_offset_),
+      j_shape(j_shape_),
+      diag(row_has_no_nd() || diag_is_first()),
+      End(false),
+      init(default_value(s))
+  { }
+
+  RowIterator(const RowIterator& rhs) : s(rhs.s), ija(rhs.ija), a(reinterpret_cast<YALE_STORAGE*>(s->src)->a), i(rhs.i), k(rhs.k), k_end(rhs.k_end), j_offset(rhs.j_offset), j_shape(rhs.j_shape), diag(rhs.diag), End(rhs.End), init(rhs.init) { }
+
+  VALUE obj() const {
+    return diag ? obj_at(s, i) : obj_at(s, k);
+  }
+
+  template <typename T>
+  T cobj() const {
+    if (typeid(T) == typeid(RubyObject)) return obj();
+    return A<T>(s)[diag ? i : k];
+  }
+
+  inline IType proper_j() const {
+    return diag ? i : ija[k];
+  }
+
+  inline IType offset_j() const {
+    return proper_j() - j_offset;
+  }
+
+  inline size_t capacity() const {
+    return reinterpret_cast<YALE_STORAGE*>(s->src)->capacity;
+  }
+
+  inline void vector_grow() {
+    YALE_STORAGE* src = reinterpret_cast<YALE_STORAGE*>(s->src);
+    nm::yale_storage::vector_grow(src);
+    ija = reinterpret_cast<IType*>(src->ija);
+    a   = src->a;
+  }
+
+  /* Returns true if an additional value is inserted, false if it goes on the diagonal */
+  bool insert(IType j, VALUE v) {
+    if (j == i) { // insert regardless on diagonal
+      reinterpret_cast<VALUE*>(a)[j] = v;
+      return false;
+
+    } else {
+      if (rb_funcall(v, rb_intern("!="), 1, init) == Qtrue) {
+        if (k >= capacity()) {
+          vector_grow();
+        }
+        reinterpret_cast<VALUE*>(a)[k] = v;
+        ija[k] = j;
+        k++;
+        return true;
+      }
+      return false;
+    }
+  }
+
+  void update_row_end() {
+    ija[i+1] = k;
+    k_end    = k;
+  }
+
+  /* Past the j_shape? */
+  inline bool end() const {
+    if (End)  return true;
+    //if (diag) return i - j_offset >= j_shape;
+    //else return k >= s->capacity || ija[k] - j_offset >= j_shape;
+    return (int)(diag ? i : ija[k]) - (int)(j_offset) >= (int)(j_shape);
+  }
+
+  inline bool row_has_no_nd() const { return ija[i] == k_end; /* k_start == k_end */  }
+  inline bool diag_is_first() const { return i < ija[ija[i]];  }
+  inline bool diag_is_last() const  { return i > ija[k_end-1]; } // only works if !row_has_no_nd()
+  inline bool k_is_last_nd() const  { return k == k_end-1;     }
+  inline bool k_is_last() const     { return k_is_last_nd() && !diag_is_last(); }
+  inline bool diag_is_ahead() const { return i > ija[k]; }
+  inline bool row_has_diag() const  { return i < s->shape[1];  }
+  inline bool diag_is_next() const  { // assumes we've already tested for diag, row_has_no_nd(), diag_is_first()
+    if (i == ija[k]+1) return true; // definite next
+    else if (k+1 < k_end && i >= ija[k+1]+1) return false; // at least one item before it
+    else return true;
+  }
+
+  RowIterator& operator++() {
+    if (diag) {                                             // we're at the diagonal
+      if (row_has_no_nd() || diag_is_last()) End = true;    //  and there are no non-diagonals (or none still to visit)
+      diag = false;
+    } else if (!row_has_diag()) {                           // row has no diagonal entries
+      if (row_has_no_nd() || k_is_last_nd()) End = true;    // row is totally empty, or we're at last entry
+      else k++;                                             // still entries to visit
+    } else { // not at diag but it exists somewhere in the row, and row has at least one nd entry
+      if (diag_is_ahead()) { // diag is ahead
+        if (k_is_last_nd()) diag = true; // diag is next and last
+        else if (diag_is_next()) {       // diag is next and not last
+          diag = true;
+          k++;
+        } else k++;                      // diag is not next
+      } else {                           // diag is past
+        if (k_is_last_nd()) End = true;  //   and we're at the end
+        else k++;                        //   and we're not at the end
+      }
+    }
+
+    return *this;
+  }
+
+
+  RowIterator operator++(int unused) {
+    RowIterator x(*this);
+    ++(*this);
+    return x;
+  }
+};
+
+
+// Helper function used only for the RETURN_SIZED_ENUMERATOR macro. Returns the length of
+// the matrix's storage.
+static VALUE nm_yale_stored_enumerator_length(VALUE nmatrix) {
+  YALE_STORAGE* s   = NM_STORAGE_YALE(nmatrix);
+  YALE_STORAGE* src = s->src == s ? s : reinterpret_cast<YALE_STORAGE*>(s->src);
+  size_t ia_size    = src->shape[0];
+  // FIXME: This needs to be corrected for slicing.
+  size_t len = std::min( s->shape[0] + s->offset[0], s->shape[1] + s->offset[1] ) + nm_yale_storage_get_size(src) -  ia_size;
+  return INT2FIX(len);
+}
+
+
+// Helper function used only for the RETURN_SIZED_ENUMERATOR macro. Returns the length of
+// the matrix's storage.
+static VALUE nm_yale_stored_nondiagonal_enumerator_length(VALUE nmatrix) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nmatrix);
+  if (s->src != s) s = reinterpret_cast<YALE_STORAGE*>(s->src);  // need to get the original storage shape
+
+  size_t ia_size = s->shape[0];
+  size_t len     = nm_yale_storage_get_size(NM_STORAGE_YALE(nmatrix)) - ia_size;
+
+  return INT2FIX(len);
+}
+
+// Helper function for diagonal length.
+static VALUE nm_yale_stored_diagonal_enumerator_length(VALUE nmatrix) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nmatrix);
+  size_t len = std::min( s->shape[0] + s->offset[0], s->shape[1] + s->offset[1] );
+  return INT2FIX(len);
+}
+
+
+// Helper function for full enumerator length.
+static VALUE nm_yale_enumerator_length(VALUE nmatrix) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nmatrix);
+  size_t len = s->shape[0] * s->shape[1];
+  return INT2FIX(len);
+}
+
+
+/*
+ * Map the stored values of a matrix in storage order.
+ */
+template <typename D>
+static VALUE map_stored(VALUE self) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(self);
+  YaleStorage<D> y(s);
+  RETURN_SIZED_ENUMERATOR(self, 0, 0, nm_yale_stored_enumerator_length);
+  YALE_STORAGE* r = y.template alloc_copy<nm::RubyObject, true>();
+  NMATRIX* m      = nm_create(nm::YALE_STORE, reinterpret_cast<STORAGE*>(r));
+  return Data_Wrap_Struct(CLASS_OF(self), nm_mark, nm_delete, m);
+}
+
+
+/*
+ * map_stored which visits the stored entries of two matrices in order.
+ */
+template <typename LD, typename RD>
+static VALUE map_merged_stored(VALUE left, VALUE right, VALUE init) {
+  nm::YaleStorage<LD> l(NM_STORAGE_YALE(left));
+  nm::YaleStorage<RD> r(NM_STORAGE_YALE(right));
+  return l.map_merged_stored(CLASS_OF(left), r, init);
+}
+
+
+/*
+ * Iterate over the stored entries in Yale (diagonal and non-diagonal non-zeros)
+ */
+template <typename DType>
+static VALUE each_stored_with_indices(VALUE nm) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nm);
+  YaleStorage<DType> y(s);
+
+  // If we don't have a block, return an enumerator.
+  RETURN_SIZED_ENUMERATOR(nm, 0, 0, nm_yale_stored_enumerator_length);
+
+  for (typename YaleStorage<DType>::const_stored_diagonal_iterator d = y.csdbegin(); d != y.csdend(); ++d) {
+    rb_yield_values(3, ~d, d.rb_i(), d.rb_j());
+  }
+
+  for (typename YaleStorage<DType>::const_row_iterator it = y.cribegin(); it != y.criend(); ++it) {
+    for (auto jt = it.ndbegin(); jt != it.ndend(); ++jt) {
+      rb_yield_values(3, ~jt, it.rb_i(), jt.rb_j());
+    }
+  }
+
+  return nm;
+}
+
+
+/*
+ * Iterate over the stored diagonal entries in Yale.
+ */
+template <typename DType>
+static VALUE stored_diagonal_each_with_indices(VALUE nm) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nm);
+  YaleStorage<DType> y(s);
+
+  // If we don't have a block, return an enumerator.
+  RETURN_SIZED_ENUMERATOR(nm, 0, 0, nm_yale_stored_diagonal_length); // FIXME: need diagonal length
+
+  for (typename YaleStorage<DType>::const_stored_diagonal_iterator d = y.csdbegin(); d != y.csdend(); ++d) {
+    rb_yield_values(3, ~d, d.rb_i(), d.rb_j());
+  }
+
+  return nm;
+}
+
+
+/*
+ * Iterate over the stored diagonal entries in Yale.
+ */
+template <typename DType>
+static VALUE stored_nondiagonal_each_with_indices(VALUE nm) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nm);
+  YaleStorage<DType> y(s);
+
+  // If we don't have a block, return an enumerator.
+  RETURN_SIZED_ENUMERATOR(nm, 0, 0, 0); // FIXME: need diagonal length
+
+  for (typename YaleStorage<DType>::const_row_iterator it = y.cribegin(); it != y.criend(); ++it) {
+    for (auto jt = it.ndbegin(); jt != it.ndend(); ++jt) {
+      rb_yield_values(3, ~jt, it.rb_i(), jt.rb_j());
+    }
+  }
+
+  return nm;
+}
+
+
+/*
+ * Iterate over the stored entries in Yale in order of i,j. Visits every diagonal entry, even if it's the default.
+ */
+template <typename DType>
+static VALUE each_ordered_stored_with_indices(VALUE nm) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nm);
+  YaleStorage<DType> y(s);
+
+  // If we don't have a block, return an enumerator.
+  RETURN_SIZED_ENUMERATOR(nm, 0, 0, nm_yale_stored_enumerator_length);
+
+  for (typename YaleStorage<DType>::const_row_iterator it = y.cribegin(); it != y.criend(); ++it) {
+    for (auto jt = it.begin(); jt != it.end(); ++jt) {
+      rb_yield_values(3, ~jt, it.rb_i(), jt.rb_j());
+    }
+  }
+
+  return nm;
+}
+
+
+template <typename DType>
+static VALUE each_with_indices(VALUE nm) {
+  YALE_STORAGE* s = NM_STORAGE_YALE(nm);
+  YaleStorage<DType> y(s);
+
+  // If we don't have a block, return an enumerator.
+  RETURN_SIZED_ENUMERATOR(nm, 0, 0, nm_yale_enumerator_length);
+
+  for (typename YaleStorage<DType>::const_iterator iter = y.cbegin(); iter != y.cend(); ++iter) {
+    rb_yield_values(3, ~iter, iter.rb_i(), iter.rb_j());
+  }
+
+  return nm;
+}
+
+
+} // end of namespace nm::yale_storage
+
+
+} // end of namespace nm.
+
+///////////////////
+// Ruby Bindings //
+///////////////////
+
+/* These bindings are mostly only for debugging Yale. They are called from Init_nmatrix. */
+
+extern "C" {
+
+void nm_init_yale_functions() {
+	/*
+	 * This module stores methods that are useful for debugging Yale matrices,
+	 * i.e. the ones with +:yale+ stype.	
+	 */
+  cNMatrix_YaleFunctions = rb_define_module_under(cNMatrix, "YaleFunctions");
+
+  rb_define_method(cNMatrix_YaleFunctions, "yale_ija", (METHOD)nm_ija, -1);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_a", (METHOD)nm_a, -1);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_size", (METHOD)nm_size, 0);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_ia", (METHOD)nm_ia, 0);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_ja", (METHOD)nm_ja, 0);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_d", (METHOD)nm_d, -1);
+  rb_define_method(cNMatrix_YaleFunctions, "yale_lu", (METHOD)nm_lu, 0);
+
+  rb_define_method(cNMatrix_YaleFunctions, "yale_nd_row", (METHOD)nm_nd_row, -1);
+
+  rb_define_const(cNMatrix_YaleFunctions, "YALE_GROWTH_CONSTANT", rb_float_new(nm::yale_storage::GROWTH_CONSTANT));
+
+  // This is so the user can easily check the IType size, mostly for debugging.
+  size_t itype_size        = sizeof(IType);
+  VALUE itype_dtype;
+  if (itype_size == sizeof(uint64_t)) {
+    itype_dtype = ID2SYM(rb_intern("int64"));
+  } else if (itype_size == sizeof(uint32_t)) {
+    itype_dtype = ID2SYM(rb_intern("int32"));
+  } else if (itype_size == sizeof(uint16_t)) {
+    itype_dtype = ID2SYM(rb_intern("int16"));
+  } else {
+    rb_raise(rb_eStandardError, "unhandled length for sizeof(IType): %lu; note that IType is probably defined as size_t", sizeof(IType));
+  }
+  rb_define_const(cNMatrix, "INDEX_DTYPE", itype_dtype);
+}
+
+
+/////////////////
+// C ACCESSORS //
+/////////////////
+
+
+/* C interface for NMatrix#each_with_indices (Yale) */
+VALUE nm_yale_each_with_indices(VALUE nmatrix) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::each_with_indices, VALUE, VALUE)
+
+  return ttable[ NM_DTYPE(nmatrix) ](nmatrix);
+}
+
+
+/* C interface for NMatrix#each_stored_with_indices (Yale) */
+VALUE nm_yale_each_stored_with_indices(VALUE nmatrix) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::each_stored_with_indices, VALUE, VALUE)
+
+  return ttable[ NM_DTYPE(nmatrix) ](nmatrix);
+}
+
+
+/* Iterate along stored diagonal (not actual diagonal!) */
+VALUE nm_yale_stored_diagonal_each_with_indices(VALUE nmatrix) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::stored_diagonal_each_with_indices, VALUE, VALUE)
+
+  return ttable[ NM_DTYPE(nmatrix) ](nmatrix);
+}
+
+/* Iterate through stored nondiagonal (not actual diagonal!) */
+VALUE nm_yale_stored_nondiagonal_each_with_indices(VALUE nmatrix) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::stored_nondiagonal_each_with_indices, VALUE, VALUE)
+
+  return ttable[ NM_DTYPE(nmatrix) ](nmatrix);
+}
+
+
+/* C interface for NMatrix#each_ordered_stored_with_indices (Yale) */
+VALUE nm_yale_each_ordered_stored_with_indices(VALUE nmatrix) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::each_ordered_stored_with_indices, VALUE, VALUE)
+
+  return ttable[ NM_DTYPE(nmatrix) ](nmatrix);
+}
+
+
+
+/*
+ * C accessor for inserting some value in a matrix (or replacing an existing cell).
+ */
+void nm_yale_storage_set(VALUE left, SLICE* slice, VALUE right) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::set, void, VALUE left, SLICE* slice, VALUE right);
+
+  ttable[NM_DTYPE(left)](left, slice, right);
+}
+
+
+/*
+ * Determine the number of non-diagonal non-zeros in a not-yet-created copy of a slice or matrix.
+ */
+static size_t yale_count_slice_copy_ndnz(const YALE_STORAGE* s, size_t* offset, size_t* shape) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::count_slice_copy_ndnz, size_t, const YALE_STORAGE*, size_t*, size_t*)
+
+  return ttable[s->dtype](s, offset, shape);
+}
+
+
+/*
+ * C accessor for yale_storage::get, which returns a slice of YALE_STORAGE object by copy
+ *
+ * Slicing-related.
+ */
+void* nm_yale_storage_get(const STORAGE* storage, SLICE* slice) {
+  YALE_STORAGE* casted_storage = (YALE_STORAGE*)storage;
+
+  if (slice->single) {
+    NAMED_DTYPE_TEMPLATE_TABLE(elem_copy_table,  nm::yale_storage::get_single, void*, YALE_STORAGE*, SLICE*)
+
+    return elem_copy_table[casted_storage->dtype](casted_storage, slice);
+  } else {
+
+    //return reinterpret_cast<void*>(nm::YaleStorage<nm::dtype_enum_T<storage->dtype>::type>(casted_storage).alloc_ref(slice));
+    NAMED_DTYPE_TEMPLATE_TABLE(ref_table, nm::yale_storage::ref, YALE_STORAGE*, YALE_STORAGE* storage, SLICE* slice)
+
+    YALE_STORAGE* ref = ref_table[casted_storage->dtype](casted_storage, slice);
+
+    NAMED_LR_DTYPE_TEMPLATE_TABLE(slice_copy_table, nm::yale_storage::slice_copy, YALE_STORAGE*, YALE_STORAGE*)
+
+    YALE_STORAGE* ns = slice_copy_table[casted_storage->dtype][casted_storage->dtype](ref);
+
+    xfree(ref);
+
+    return ns;
+  }
+}
+
+/*
+ * C accessor for yale_storage::vector_insert
+ */
+static char nm_yale_storage_vector_insert(YALE_STORAGE* s, size_t pos, size_t* js, void* vals, size_t n, bool struct_only, nm::dtype_t dtype) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::vector_insert, char, YALE_STORAGE*, size_t, size_t*, void*, size_t, bool);
+
+  return ttable[dtype](s, pos, js, vals, n, struct_only);
+}
+
+/*
+ * C accessor for yale_storage::increment_ia_after, typically called after ::vector_insert
+ */
+static void nm_yale_storage_increment_ia_after(YALE_STORAGE* s, size_t ija_size, size_t i, long n) {
+  nm::yale_storage::increment_ia_after(s, ija_size, i, n);
+}
+
+
+/*
+ * C accessor for yale_storage::ref, which returns either a pointer to the correct location in a YALE_STORAGE object
+ * for some set of coordinates, or a pointer to a single element.
+ */
+void* nm_yale_storage_ref(const STORAGE* storage, SLICE* slice) {
+  YALE_STORAGE* casted_storage = (YALE_STORAGE*)storage;
+
+  if (slice->single) {
+    //return reinterpret_cast<void*>(nm::YaleStorage<nm::dtype_enum_T<storage->dtype>::type>(casted_storage).get_single_p(slice));
+    NAMED_DTYPE_TEMPLATE_TABLE(elem_copy_table,  nm::yale_storage::get_single, void*, YALE_STORAGE*, SLICE*)
+    return elem_copy_table[casted_storage->dtype](casted_storage, slice);
+  } else {
+    //return reinterpret_cast<void*>(nm::YaleStorage<nm::dtype_enum_T<storage->dtype>::type>(casted_storage).alloc_ref(slice));
+    NAMED_DTYPE_TEMPLATE_TABLE(ref_table, nm::yale_storage::ref, YALE_STORAGE*, YALE_STORAGE* storage, SLICE* slice)
+    return reinterpret_cast<void*>(ref_table[casted_storage->dtype](casted_storage, slice));
+
+  }
+}
+
+
+/*
+ * C accessor for determining whether two YALE_STORAGE objects have the same contents.
+ */
+bool nm_yale_storage_eqeq(const STORAGE* left, const STORAGE* right) {
+  NAMED_LR_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::eqeq, bool, const YALE_STORAGE* left, const YALE_STORAGE* right);
+
+  const YALE_STORAGE* casted_left = reinterpret_cast<const YALE_STORAGE*>(left);
+
+  return ttable[casted_left->dtype][right->dtype](casted_left, (const YALE_STORAGE*)right);
+}
+
+
+/*
+ * Copy constructor for changing dtypes. (C accessor)
+ */
+STORAGE* nm_yale_storage_cast_copy(const STORAGE* rhs, nm::dtype_t new_dtype, void* dummy) {
+  NAMED_LR_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::cast_copy, YALE_STORAGE*, const YALE_STORAGE* rhs);
+
+  const YALE_STORAGE* casted_rhs = reinterpret_cast<const YALE_STORAGE*>(rhs);
+  //return reinterpret_cast<STORAGE*>(nm::YaleStorage<nm::dtype_enum_T< rhs->dtype >::type>(rhs).alloc_copy<nm::dtype_enum_T< new_dtype >::type>());
+  return (STORAGE*)ttable[new_dtype][casted_rhs->dtype](casted_rhs);
+}
+
+
+/*
+ * Returns size of Yale storage as a size_t (no matter what the itype is). (C accessor)
+ */
+size_t nm_yale_storage_get_size(const YALE_STORAGE* storage) {
+  return nm::yale_storage::get_size(storage);
+}
+
+
+
+/*
+ * Return a pointer to the matrix's default value entry.
+ */
+static void* default_value_ptr(const YALE_STORAGE* s) {
+  return reinterpret_cast<void*>(reinterpret_cast<char*>(((YALE_STORAGE*)(s->src))->a) + (((YALE_STORAGE*)(s->src))->shape[0] * DTYPE_SIZES[s->dtype]));
+}
+
+/*
+ * Return the Ruby object at a given location in storage.
+ */
+static VALUE obj_at(YALE_STORAGE* s, size_t k) {
+  if (s->dtype == nm::RUBYOBJ)  return reinterpret_cast<VALUE*>(((YALE_STORAGE*)(s->src))->a)[k];
+  else  return rubyobj_from_cval(reinterpret_cast<void*>(reinterpret_cast<char*>(((YALE_STORAGE*)(s->src))->a) + k * DTYPE_SIZES[s->dtype]), s->dtype).rval;
+}
+
+
+/*
+ * Return the matrix's default value as a Ruby VALUE.
+ */
+static VALUE default_value(const YALE_STORAGE* s) {
+  if (s->dtype == nm::RUBYOBJ) return *reinterpret_cast<VALUE*>(default_value_ptr(s));
+  else return rubyobj_from_cval(default_value_ptr(s), s->dtype).rval;
+}
+
+
+/*
+ * Check to see if a default value is some form of zero. Easy for non-Ruby object matrices, which should always be 0.
+ */
+static bool default_value_is_numeric_zero(const YALE_STORAGE* s) {
+  return rb_funcall(default_value(s), rb_intern("=="), 1, INT2FIX(0)) == Qtrue;
+}
+
+
+
+/*
+ * Transposing copy constructor.
+ */
+STORAGE* nm_yale_storage_copy_transposed(const STORAGE* rhs_base) {
+  YALE_STORAGE* rhs = (YALE_STORAGE*)rhs_base;
+  NAMED_DTYPE_TEMPLATE_TABLE(transp, nm::yale_storage::copy_transposed, YALE_STORAGE*, YALE_STORAGE*)
+  return (STORAGE*)(transp[rhs->dtype](rhs));
+}
+
+/*
+ * C accessor for multiplying two YALE_STORAGE matrices, which have already been casted to the same dtype.
+ *
+ * FIXME: There should be some mathematical way to determine the worst-case IType based on the input ITypes. Right now
+ * it just uses the default.
+ */
+STORAGE* nm_yale_storage_matrix_multiply(const STORAGE_PAIR& casted_storage, size_t* resulting_shape, bool vector) {
+  DTYPE_TEMPLATE_TABLE(nm::yale_storage::matrix_multiply, STORAGE*, const STORAGE_PAIR& casted_storage, size_t* resulting_shape, bool vector);
+
+  YALE_STORAGE* left = reinterpret_cast<YALE_STORAGE*>(casted_storage.left);
+  YALE_STORAGE* right = reinterpret_cast<YALE_STORAGE*>(casted_storage.right);
+
+  if (!default_value_is_numeric_zero(left) || !default_value_is_numeric_zero(right)) {
+    rb_raise(rb_eNotImpError, "matrix default value must be some form of zero (not false or nil) for multiplication");
+    return NULL;
+  }
+
+  return ttable[left->dtype](casted_storage, resulting_shape, vector);
+}
+
+
+///////////////
+// Lifecycle //
+///////////////
+
+/*
+ * C accessor function for creating a YALE_STORAGE object. Prior to calling this function, you MUST
+ * allocate shape (should be size_t * 2) -- don't use use a regular size_t array!
+ *
+ * For this type, dim must always be 2. The final argument is the initial capacity with which to
+ * create the storage.
+ */
+
+YALE_STORAGE* nm_yale_storage_create(nm::dtype_t dtype, size_t* shape, size_t dim, size_t init_capacity) {
+  if (dim != 2) {
+    rb_raise(nm_eStorageTypeError, "yale supports only 2-dimensional matrices");
+  }
+  DTYPE_OBJECT_STATIC_TABLE(nm::YaleStorage, create, YALE_STORAGE*, size_t* shape, size_t init_capacity)
+  return ttable[dtype](shape, init_capacity);
+}
+
+/*
+ * Destructor for yale storage (C-accessible).
+ */
+void nm_yale_storage_delete(STORAGE* s) {
+  if (s) {
+    YALE_STORAGE* storage = (YALE_STORAGE*)s;
+    if (storage->count-- == 1) {
+      xfree(storage->shape);
+      xfree(storage->offset);
+      xfree(storage->ija);
+      xfree(storage->a);
+      xfree(storage);
+    }
+  }
+}
+
+/*
+ * Destructor for the yale storage ref
+ */
+void nm_yale_storage_delete_ref(STORAGE* s) {
+  if (s) {
+    YALE_STORAGE* storage = (YALE_STORAGE*)s;
+    nm_yale_storage_delete( reinterpret_cast<STORAGE*>(storage->src) );
+    xfree(storage->shape);
+    xfree(storage->offset);
+    xfree(s);
+  }
+}
+
+/*
+ * C accessor for yale_storage::init, a templated function.
+ *
+ * Initializes the IJA vector of the YALE_STORAGE matrix.
+ */
+void nm_yale_storage_init(YALE_STORAGE* s, void* init_val) {
+  DTYPE_TEMPLATE_TABLE(nm::yale_storage::init, void, YALE_STORAGE*, void*);
+
+  ttable[s->dtype](s, init_val);
+}
+
+
+/*
+ * Ruby GC mark function for YALE_STORAGE. C accessible.
+ */
+void nm_yale_storage_mark(STORAGE* storage_base) {
+  YALE_STORAGE* storage = (YALE_STORAGE*)storage_base;
+  size_t i;
+
+  if (storage && storage->dtype == nm::RUBYOBJ) {
+
+    VALUE* a = (VALUE*)(storage->a);
+    rb_gc_mark_locations(a, a + storage->capacity * sizeof(VALUE));
+  }
+}
+
+
+/*
+ * Allocates and initializes the basic struct (but not the IJA or A vectors).
+ *
+ * This function is ONLY used when creating from old yale.
+ */
+static YALE_STORAGE* alloc(nm::dtype_t dtype, size_t* shape, size_t dim) {
+  YALE_STORAGE* s;
+
+  s = ALLOC( YALE_STORAGE );
+
+  s->ndnz        = 0;
+  s->dtype       = dtype;
+  s->shape       = shape;
+  s->offset      = ALLOC_N(size_t, dim);
+  for (size_t i = 0; i < dim; ++i)
+    s->offset[i] = 0;
+  s->dim         = dim;
+  s->src         = reinterpret_cast<STORAGE*>(s);
+  s->count       = 1;
+
+  return s;
+}
+
+YALE_STORAGE* nm_yale_storage_create_from_old_yale(nm::dtype_t dtype, size_t* shape, char* ia, char* ja, char* a, nm::dtype_t from_dtype) {
+  NAMED_LR_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::create_from_old_yale, YALE_STORAGE*, nm::dtype_t dtype, size_t* shape, char* r_ia, char* r_ja, char* r_a);
+
+  return ttable[dtype][from_dtype](dtype, shape, ia, ja, a);
+
+}
+
+//////////////////////////////////////////////
+// YALE-SPECIFIC FUNCTIONS (RUBY ACCESSORS) //
+//////////////////////////////////////////////
+
+/*
+ * call-seq:
+ *     yale_size -> Integer
+ *
+ * Get the size of a Yale matrix (the number of elements actually stored).
+ *
+ * For capacity (the maximum number of elements that can be stored without a resize), use capacity instead.
+ */
+static VALUE nm_size(VALUE self) {
+  YALE_STORAGE* s = (YALE_STORAGE*)(NM_SRC(self));
+  return INT2FIX(nm::yale_storage::IJA(s)[s->shape[0]]);
+}
+
+
+/*
+ * call-seq:
+ *     yale_a -> Array
+ *     yale_d(index) -> ...
+ *
+ * Get the A array of a Yale matrix (which stores the diagonal and the LU portions of the matrix).
+ */
+static VALUE nm_a(int argc, VALUE* argv, VALUE self) {
+  VALUE idx;
+  rb_scan_args(argc, argv, "01", &idx);
+
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+  size_t size = nm_yale_storage_get_size(s);
+
+  if (idx == Qnil) {
+    VALUE* vals = ALLOCA_N(VALUE, size);
+
+    if (NM_DTYPE(self) == nm::RUBYOBJ) {
+      for (size_t i = 0; i < size; ++i) {
+        vals[i] = reinterpret_cast<VALUE*>(s->a)[i];
+      }
+    } else {
+      for (size_t i = 0; i < size; ++i) {
+        vals[i] = rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype]*i, s->dtype).rval;
+      }
+    }
+    VALUE ary = rb_ary_new4(size, vals);
+
+    for (size_t i = size; i < s->capacity; ++i)
+      rb_ary_push(ary, Qnil);
+
+    return ary;
+  } else {
+    size_t index = FIX2INT(idx);
+    if (index >= size) rb_raise(rb_eRangeError, "out of range");
+
+    return rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype] * index, s->dtype).rval;
+  }
+}
+
+
+/*
+ * call-seq:
+ *     yale_d -> Array
+ *     yale_d(index) -> ...
+ *
+ * Get the diagonal ("D") portion of the A array of a Yale matrix.
+ */
+static VALUE nm_d(int argc, VALUE* argv, VALUE self) {
+  VALUE idx;
+  rb_scan_args(argc, argv, "01", &idx);
+
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+
+  if (idx == Qnil) {
+    VALUE* vals = ALLOCA_N(VALUE, s->shape[0]);
+
+    if (NM_DTYPE(self) == nm::RUBYOBJ) {
+      for (size_t i = 0; i < s->shape[0]; ++i) {
+        vals[i] = reinterpret_cast<VALUE*>(s->a)[i];
+      }
+    } else {
+      for (size_t i = 0; i < s->shape[0]; ++i) {
+        vals[i] = rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype]*i, s->dtype).rval;
+      }
+    }
+
+    return rb_ary_new4(s->shape[0], vals);
+  } else {
+    size_t index = FIX2INT(idx);
+    if (index >= s->shape[0]) rb_raise(rb_eRangeError, "out of range");
+
+    return rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype] * index, s->dtype).rval;
+  }
+}
+
+/*
+ * call-seq:
+ *     yale_lu -> Array
+ *
+ * Get the non-diagonal ("LU") portion of the A array of a Yale matrix.
+ */
+static VALUE nm_lu(VALUE self) {
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+
+  size_t size = nm_yale_storage_get_size(s);
+
+  VALUE* vals = ALLOCA_N(VALUE, size - s->shape[0] - 1);
+
+  if (NM_DTYPE(self) == nm::RUBYOBJ) {
+    for (size_t i = 0; i < size - s->shape[0] - 1; ++i) {
+      vals[i] = reinterpret_cast<VALUE*>(s->a)[s->shape[0] + 1 + i];
+    }
+  } else {
+    for (size_t i = 0; i < size - s->shape[0] - 1; ++i) {
+      vals[i] = rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype]*(s->shape[0] + 1 + i), s->dtype).rval;
+    }
+  }
+
+  VALUE ary = rb_ary_new4(size - s->shape[0] - 1, vals);
+
+  for (size_t i = size; i < s->capacity; ++i)
+    rb_ary_push(ary, Qnil);
+
+  return ary;
+}
+
+/*
+ * call-seq:
+ *     yale_ia -> Array
+ *
+ * Get the IA portion of the IJA array of a Yale matrix. This gives the start and end positions of rows in the
+ * JA and LU portions of the IJA and A arrays, respectively.
+ */
+static VALUE nm_ia(VALUE self) {
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+
+  VALUE* vals = ALLOCA_N(VALUE, s->shape[0] + 1);
+
+  for (size_t i = 0; i < s->shape[0] + 1; ++i) {
+    vals[i] = INT2FIX(s->ija[i]);
+  }
+
+  return rb_ary_new4(s->shape[0]+1, vals);
+}
+
+/*
+ * call-seq:
+ *     yale_ja -> Array
+ *
+ * Get the JA portion of the IJA array of a Yale matrix. This gives the column indices for entries in corresponding
+ * positions in the LU portion of the A array.
+ */
+static VALUE nm_ja(VALUE self) {
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+
+  size_t size = nm_yale_storage_get_size(s);
+
+  VALUE* vals = ALLOCA_N(VALUE, size - s->shape[0] - 1);
+
+  for (size_t i = 0; i < size - s->shape[0] - 1; ++i) {
+    vals[i] = INT2FIX(s->ija[s->shape[0] + 1 + i]);
+  }
+
+  VALUE ary = rb_ary_new4(size - s->shape[0] - 1, vals);
+
+  for (size_t i = size; i < s->capacity; ++i)
+    rb_ary_push(ary, Qnil);
+
+  return ary;
+}
+
+/*
+ * call-seq:
+ *     yale_ija -> Array
+ *     yale_ija(index) -> ...
+ *
+ * Get the IJA array of a Yale matrix (or a component of the IJA array).
+ */
+static VALUE nm_ija(int argc, VALUE* argv, VALUE self) {
+  VALUE idx;
+  rb_scan_args(argc, argv, "01", &idx);
+
+  YALE_STORAGE* s = reinterpret_cast<YALE_STORAGE*>(NM_SRC(self));
+  size_t size = nm_yale_storage_get_size(s);
+
+  if (idx == Qnil) {
+
+    VALUE* vals = ALLOCA_N(VALUE, size);
+
+    for (size_t i = 0; i < size; ++i) {
+      vals[i] = INT2FIX(s->ija[i]);
+    }
+
+   VALUE ary = rb_ary_new4(size, vals);
+
+    for (size_t i = size; i < s->capacity; ++i)
+      rb_ary_push(ary, Qnil);
+
+    return ary;
+
+  } else {
+    size_t index = FIX2INT(idx);
+    if (index >= size) rb_raise(rb_eRangeError, "out of range");
+
+    return INT2FIX(s->ija[index]);
+  }
+}
+
+
+/*
+ * call-seq:
+ *     yale_nd_row -> ...
+ *
+ * This function gets the non-diagonal contents of a Yale matrix row.
+ * The first argument should be the row index. The optional second argument may be :hash or :keys, but defaults
+ * to :hash. If :keys is given, it will only return the Hash keys (the column indices).
+ *
+ * This function is meant to accomplish its purpose as efficiently as possible. It does not check for appropriate
+ * range.
+ */
+static VALUE nm_nd_row(int argc, VALUE* argv, VALUE self) {
+  if (NM_SRC(self) != NM_STORAGE(self))
+    rb_raise(rb_eNotImpError, "must be called on a real matrix and not a slice");
+
+  VALUE i_, as;
+  rb_scan_args(argc, argv, "11", &i_, &as);
+
+  bool keys = false;
+  if (as != Qnil && rb_to_id(as) != nm_rb_hash) keys = true;
+
+  size_t i = FIX2INT(i_);
+
+  YALE_STORAGE* s   = NM_STORAGE_YALE(self);
+  nm::dtype_t dtype = NM_DTYPE(self);
+
+  size_t pos = s->ija[i];
+  size_t nextpos = s->ija[i+1];
+  size_t diff = nextpos - pos;
+
+  VALUE ret;
+  if (keys) {
+    ret = rb_ary_new3(diff);
+
+    for (size_t idx = pos; idx < nextpos; ++idx) {
+      rb_ary_store(ret, idx - pos, INT2FIX(s->ija[idx]));
+    }
+
+  } else {
+    ret = rb_hash_new();
+
+    for (size_t idx = pos; idx < nextpos; ++idx) {
+      rb_hash_aset(ret, INT2FIX(s->ija[idx]), rubyobj_from_cval((char*)(s->a) + DTYPE_SIZES[s->dtype]*idx, s->dtype).rval);
+    }
+  }
+
+  return ret;
+}
+
+/*
+ * call-seq:
+ *     yale_vector_set(i, column_index_array, cell_contents_array, pos) -> Fixnum
+ *
+ * Insert at position pos an array of non-diagonal elements with column indices given. Note that the column indices and values
+ * must be storage-contiguous -- that is, you can't insert them around existing elements in some row, only amid some
+ * elements in some row. You *can* insert them around a diagonal element, since this is stored separately. This function
+ * may not be used for the insertion of diagonal elements in most cases, as these are already present in the data
+ * structure and are typically modified by replacement rather than insertion.
+ *
+ * The last argument, pos, may be nil if you want to insert at the beginning of a row. Otherwise it needs to be provided.
+ * Don't expect this function to know the difference. It really does very little checking, because its goal is to make
+ * multiple contiguous insertion as quick as possible.
+ *
+ * You should also not attempt to insert values which are the default (0). These are not supposed to be stored, and may
+ * lead to undefined behavior.
+ *
+ * Example:
+ *    m.yale_vector_set(3, [0,3,4], [1,1,1], 15)
+ *
+ * The example above inserts the values 1, 1, and 1 in columns 0, 3, and 4, assumed to be located at position 15 (which
+ * corresponds to row 3).
+ *
+ * Example:
+ *    next = m.yale_vector_set(3, [0,3,4], [1,1,1])
+ *
+ * This example determines that i=3 is at position 15 automatically. The value returned, next, is the position where the
+ * next value(s) should be inserted.
+ */
+VALUE nm_vector_set(int argc, VALUE* argv, VALUE self) { //, VALUE i_, VALUE jv, VALUE vv, VALUE pos_) {
+
+  if (NM_SRC(self) != NM_STORAGE(self))
+    rb_raise(rb_eNotImpError, "must be called on a real matrix and not a slice");
+
+  // i, jv, vv are mandatory; pos is optional; thus "31"
+  VALUE i_, jv, vv, pos_;
+  rb_scan_args(argc, argv, "31", &i_, &jv, &vv, &pos_);
+
+  size_t len   = RARRAY_LEN(jv); // need length in order to read the arrays in
+  size_t vvlen = RARRAY_LEN(vv);
+
+  if (len != vvlen)
+    rb_raise(rb_eArgError, "lengths must match between j array (%d) and value array (%d)", len, vvlen);
+
+  YALE_STORAGE* s   = NM_STORAGE_YALE(self);
+  nm::dtype_t dtype = NM_DTYPE(self);
+
+  size_t i   = FIX2INT(i_);    // get the row
+  size_t pos = s->ija[i];
+
+  // Allocate the j array and the values array
+  size_t* j  = ALLOCA_N(size_t, len);
+  void* vals = ALLOCA_N(char, DTYPE_SIZES[dtype] * len);
+
+  // Copy array contents
+  for (size_t idx = 0; idx < len; ++idx) {
+    j[idx] = FIX2INT(rb_ary_entry(jv, idx));
+    rubyval_to_cval(rb_ary_entry(vv, idx), dtype, (char*)vals + idx * DTYPE_SIZES[dtype]);
+  }
+
+  nm_yale_storage_vector_insert(s, pos, j, vals, len, false, dtype);
+  nm_yale_storage_increment_ia_after(s, s->shape[0], i, len);
+  s->ndnz += len;
+
+  // Return the updated position
+  pos += len;
+  return INT2FIX(pos);
+}
+
+
+
+
+/*
+ * call-seq:
+ *     __yale_default_value__ -> ...
+ *
+ * Get the default_value property from a yale matrix.
+ */
+VALUE nm_yale_default_value(VALUE self) {
+  return default_value(NM_STORAGE_YALE(self));
+}
+
+
+/*
+ * call-seq:
+ *     __yale_map_merged_stored__(right) -> Enumerator
+ *
+ * A map operation on two Yale matrices which only iterates across the stored indices.
+ */
+VALUE nm_yale_map_merged_stored(VALUE left, VALUE right, VALUE init) {
+  NAMED_LR_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::map_merged_stored, VALUE, VALUE, VALUE, VALUE)
+  return ttable[NM_DTYPE(left)][NM_DTYPE(right)](left, right, init);
+  //return nm::yale_storage::map_merged_stored(left, right, init);
+}
+
+
+/*
+ * call-seq:
+ *     __yale_map_stored__ -> Enumerator
+ *
+ * A map operation on two Yale matrices which only iterates across the stored indices.
+ */
+VALUE nm_yale_map_stored(VALUE self) {
+  NAMED_DTYPE_TEMPLATE_TABLE(ttable, nm::yale_storage::map_stored, VALUE, VALUE)
+  return ttable[NM_DTYPE(self)](self);
+}
+
+} // end of extern "C" block