Support for sparse/dense vectors in the python extension

- SparseVector, DenseVector
- improved Lattice
- Lattice translation
- Hypergraph reweighting, pruning

1 files changed, 101 insertions, 0 deletions

diff --git a/python/src/vectors.pxi b/python/src/vectors.pxi
new file mode 100644
index 00000000..74d3a0bd
--- /dev/null
+++ b/python/src/vectors.pxi
@@ -0,0 +1,101 @@
+from cython.operator cimport preincrement as pinc
+
+cdef class DenseVector:
+    cdef vector[weight_t]* vector
+
+    def __getitem__(self, char* fname):
+        cdef unsigned fid = FDConvert(fname)
+        if fid <= self.vector.size():
+            return self.vector[0][fid]
+        raise KeyError(fname)
+    
+    def __setitem__(self, char* fname, float value):
+        cdef unsigned fid = FDConvert(<char *>fname)
+        if self.vector.size() <= fid:
+            self.vector.resize(fid + 1)
+        self.vector[0][fid] = value
+
+    def __iter__(self):
+        cdef unsigned fid
+        for fid in range(1, self.vector.size()):
+            yield FDConvert(fid).c_str(), self.vector[0][fid]
+
+    def dot(self, SparseVector other):
+        return other.dot(self)
+
+    def tosparse(self):
+        cdef SparseVector sparse = SparseVector()
+        sparse.vector = new FastSparseVector[weight_t]()
+        InitSparseVector(self.vector[0], sparse.vector)
+        return sparse
+
+cdef class SparseVector:
+    cdef FastSparseVector[weight_t]* vector
+
+    def __getitem__(self, char* fname):
+        cdef unsigned fid = FDConvert(fname)
+        return self.vector.value(fid)
+    
+    def __setitem__(self, char* fname, float value):
+        cdef unsigned fid = FDConvert(<char *>fname)
+        self.vector.set_value(fid, value)
+
+    def __iter__(self):
+        cdef FastSparseVector[weight_t].const_iterator* it = new FastSparseVector[weight_t].const_iterator(self.vector[0], False)
+        cdef str fname
+        for i in range(self.vector.size()):
+            fname = FDConvert(it[0].ptr().first).c_str()
+            yield (fname, it[0].ptr().second)
+            pinc(it[0])
+
+    def dot(self, other):
+        if isinstance(other, DenseVector):
+            return self.vector.dot((<DenseVector> other).vector[0])
+        elif isinstance(other, SparseVector):
+            return self.vector.dot((<SparseVector> other).vector[0])
+        raise ValueError('cannot take the dot product of %s and SparseVector' % type(other))
+
+    def todense(self):
+        cdef DenseVector dense = DenseVector()
+        dense.vector = new vector[weight_t]()
+        self.vector.init_vector(dense.vector)
+        return dense
+    
+    def __richcmp__(SparseVector self, SparseVector other, int op):
+        if op == 2: # ==
+            return self.vector[0] == other.vector[0]
+        elif op == 3: # !=
+            return not (self == other)
+        raise NotImplemented('comparison not implemented for SparseVector')
+
+    def __len__(self):
+        return self.vector.size()
+
+    def __contains__(self, char* fname):
+        return self.vector.nonzero(FDConvert(fname))
+    
+    def __iadd__(SparseVector self, SparseVector other):
+        self.vector[0] += other.vector[0]
+        return self
+
+    def __isub__(SparseVector self, SparseVector other):
+        self.vector[0] -= other.vector[0]
+        return self
+
+    def __imul__(SparseVector self, float scalar):
+        self.vector[0] *= scalar
+        return self
+
+    def __idiv__(SparseVector self, float scalar):
+        self.vector[0] /= scalar
+        return self
+
+    def __add__(SparseVector self, SparseVector other):
+        cdef SparseVector result = SparseVector()
+        result.vector = new FastSparseVector[weight_t](self.vector[0] + other.vector[0])
+        return result
+
+    def __sub__(SparseVector self, SparseVector other):
+        cdef SparseVector result = SparseVector()
+        result.vector = new FastSparseVector[weight_t](self.vector[0] - other.vector[0])
+        return result