class LLMatFrobeniusNormBenchmark(benchmark.Benchmark):
label = "Generalize norm with 100 elements and size = 1,000 for a symmetrical matrix"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T, is_symmetric=True)
construct_sym_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat_sym(self.size, self.size, self.nbr_elements)
construct_sym_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# assert self.A_c.nnz == self.A_p.nnz
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.generalize()
return
def test_cysparse(self):
self.A_c.generalize()
return
class LLMatColScaleBenchmark(benchmark.Benchmark):
label = "Simple col_scale with 100 elements and size = 1,000"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.col_scale(self.v)
return
def test_cysparse(self):
self.A_c.col_scale(self.v)
return
class LLMatColScaleBenchmark(benchmark.Benchmark):
label = "Simple col_scale with 100 elements and size = 1,000"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.col_scale(self.v)
return
def test_cysparse(self):
self.A_c.col_scale(self.v)
return
class LLMatFrobeniusNormBenchmark(benchmark.Benchmark):
label = "Generalize norm with 100 elements and size = 1,000 for a symmetrical matrix"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T,
is_symmetric=True)
construct_sym_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat_sym(self.size, self.size, self.nbr_elements)
construct_sym_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# assert self.A_c.nnz == self.A_p.nnz
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.generalize()
return
def test_cysparse(self):
self.A_c.generalize()
return
class LLMatShiftBenchmark(benchmark.Benchmark):
label = "Simple shift with 100 elements and size = 1,000 (sigma = 10.47)"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.sigma = 10.47
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_c2 = self.A_c.copy()
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.A_p2 = self.A_p.copy()
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.shift(self.sigma, self.A_p2)
return
def test_cysparse(self):
self.A_c.shift(self.sigma, self.A_c2)
return
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.non_contiguity = 10
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size,
self.nbr_elements)
self.CSC_c = self.A_c.to_csc()
self.CSC_s = self.A_s.tocsc()
self.v = np.arange(0,
self.size * self.non_contiguity,
dtype=np.float64)
def factorize(self):
u"""Factorize matrix A as limited-memory LDLᵀ.
:returns:
:L: L as a llmat matrix
:d: as a Numpy array
"""
super(CySparseLLDLSolver, self).factorize()
nnz = len(self.lvals)
row = np.empty(nnz, dtype=np.int64)
col = np.empty(nnz, dtype=np.int64)
val = np.empty(nnz, dtype=np.float64)
elem = 0
for j in xrange(len(self.colptr) - 1):
for k in xrange(self.colptr[j], self.colptr[j + 1]):
row[elem] = self.rowind[k]
col[elem] = j
val[elem] = self.lvals[k]
elem += 1
L = LLSparseMatrix(size=self.n,
itype=INT64_T,
dtype=FLOAT64_T)
L.put_triplet(row, col, val)
return (L, self.d)
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSR_c = self.A_c.to_csr()
self.CSR_s = self.A_s.tocsr()
self.B_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.B_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.B_c)
self.list_of_matrices.append(self.B_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSC_c = self.B_c.to_csc()
self.CSC_s = self.B_s.tocsc()
self.v = np.arange(0, self.size, dtype=np.float64)
class LLMatRowScaleStridedBenchmark(benchmark.Benchmark):
label = "Simple row_scale with 100 elements and size = 1,000 and a strided NumPy vector (stride = 10)"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.stride = 10
self.v = np.arange(0, self.size * self.stride, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.row_scale(self.v[::self.stride])
return
def test_cysparse(self):
self.A_c.row_scale(self.v[::self.stride])
return
class LLMatMatVecBenchmark(benchmark.Benchmark):
label = "CSR * CSC * v with 1000 elements and size = 10,000"
each = 100
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size,
self.nbr_elements)
self.CSR_c = self.A_c.to_csr()
self.CSR_s = self.A_s.tocsr()
self.B_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
self.B_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.B_c)
self.list_of_matrices.append(self.B_s)
construct_random_matrices(self.list_of_matrices, self.size,
self.nbr_elements)
self.CSC_c = self.B_c.to_csc()
self.CSC_s = self.B_s.tocsc()
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# assert self.w_c[i] == self.w_p[i]
# assert self.w_c[i] == self.w_s[i]
def test_cysparse(self):
self.w_c = self.CSR_c * self.CSC_c * self.v
return
def test_scipy_sparse(self):
self.w_s = self.CSR_s * self.CSC_s * self.v
return
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def jac(self, *args, **kwargs):
"""Evaluate constraints Jacobian at x."""
vals, rows, cols = super(CySparseNLPModel, self).jac(*args, **kwargs)
J = LLSparseMatrix(nrow=self.ncon,
ncol=self.nvar,
size_hint=vals.size,
store_symmetric=False,
itype=types.INT64_T,
dtype=types.FLOAT64_T)
J.put_triplet(rows, cols, vals)
return J
class LLMatMatVecBenchmark(benchmark.Benchmark):
label = "CSR * CSC * v with 1000 elements and size = 10,000"
each = 100
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSR_c = self.A_c.to_csr()
self.CSR_s = self.A_s.tocsr()
self.B_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.B_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.B_c)
self.list_of_matrices.append(self.B_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSC_c = self.B_c.to_csc()
self.CSC_s = self.B_s.tocsc()
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# assert self.w_c[i] == self.w_p[i]
# assert self.w_c[i] == self.w_s[i]
def test_cysparse(self):
self.w_c = self.CSR_c * self.CSC_c * self.v
return
def test_scipy_sparse(self):
self.w_s = self.CSR_s * self.CSC_s * self.v
return
class LLMatMatVecBenchmark(benchmark.Benchmark):
label = "matvec with 1000 elements and size = 10,000"
each = 100
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_p)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# assert self.w_c[i] == self.w_p[i]
# assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.w_p = np.empty(self.size, dtype=np.float64)
self.A_p.matvec(self.v, self.w_p)
return
def test_cysparse(self):
self.w_c = self.A_c * self.v
return
def test_cysparse2(self):
self.A_c.matvec(self.v)
return
def test_scipy_sparse(self):
self.w_s = self.A_s * self.v
return
def test_scipy_sparse2(self):
self.A_s._mul_vector(self.v)
class LLMatMatVecBenchmark(benchmark.Benchmark):
label = "matvec with 1000 elements and size = 10,000"
each = 100
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_p)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size,
self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# assert self.w_c[i] == self.w_p[i]
# assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.w_p = np.empty(self.size, dtype=np.float64)
self.A_p.matvec(self.v, self.w_p)
return
def test_cysparse(self):
self.w_c = self.A_c * self.v
return
def test_cysparse2(self):
self.A_c.matvec(self.v)
return
def test_scipy_sparse(self):
self.w_s = self.A_s * self.v
return
def test_scipy_sparse2(self):
self.A_s._mul_vector(self.v)
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.stride = 10
self.v = np.arange(0, self.size * self.stride, dtype=np.float64)
def hess(self, *args, **kwargs):
"""Evaluate Lagrangian Hessian at (x, z).
Note that `rows`, `cols` and `vals` must represent a LOWER triangular
sparse matrix in the coordinate format (COO).
"""
vals, rows, cols = super(CySparseNLPModel, self).hess(*args, **kwargs)
H = LLSparseMatrix(size=self.nvar,
size_hint=vals.size,
store_symmetric=True,
itype=types.INT64_T,
dtype=types.FLOAT64_T)
H.put_triplet(rows, cols, vals)
return H
class LLMatPutTripletBenchmark(benchmark.Benchmark):
label = "Simple put_triplet with 100 elements, size = 1,000 and put_size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.put_size = 1000
assert self.put_size <= self.size
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.A_sppy = None
self.id1 = np.arange(0, self.put_size, dtype=np.int32)
self.id2 = np.full(self.put_size, 37, dtype=np.int32)
self.b = np.arange(0, self.put_size, dtype=np.float64)
def eachSetUp(self):
self.A_sppy = csarray((self.size, self.size),
dtype=np.float64,
storagetype='row')
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i, j] == self.A_p[i, j]
def test_pysparse(self):
self.A_p.put(self.b, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.put_triplet(self.id1, self.id2, self.b)
return
def test_sppy(self):
self.A_sppy.put(self.b, self.id1, self.id2, init=True)
class LLMatInfiniteNormBenchmark(benchmark.Benchmark):
label = "Infinite norm with 100 elements and size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def tearDown(self):
assert self.p_norm_inf == self.c_norm_inf
#assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.p_norm_inf = self.A_p.norm('inf')
return
def test_cysparse(self):
self.c_norm_inf = self.A_c.norm('inf')
return
class LLMatToCSRBenchmark(benchmark.Benchmark):
label = "to_csr() with 100 elements and size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.csr_c[i, j] == self.A_c[i, j]
def test_pysparse(self):
self.csr_p = self.A_p.to_csr()
return
def test_cysparse(self):
self.csr_c = self.A_c.to_csr()
return
class LLMatMatVecBenchmark_2(LLMatMatVecBenchmark):
label = "matvec with 10,000 elements and size = 100,000"
each = 100
def setUp(self):
self.nbr_elements = 10000
self.size = 100000
self.non_contiguity = 10
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSC_c = self.A_c.to_csc()
self.CSC_s = self.A_s.tocsc()
self.v = np.arange(0, self.size * self.non_contiguity, dtype=np.float64)
class LLMatFrobeniusNormBenchmark(benchmark.Benchmark):
label = "Frobenius norm with 100 elements and size = 1,000 for a symmetrical matrix"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T, store_symmetric=True)
construct_sym_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat_sym(self.size, self.size, self.nbr_elements)
construct_sym_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def tearDown(self):
assert self.p_norm == self.c_norm
#assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.p_norm = self.A_p.norm('fro')
return
def test_cysparse(self):
self.c_norm = self.A_c.norm('frob')
return
class LLMatMatVecBenchmark_4(LLMatMatVecBenchmark):
label = "matvec with 5000 elements and size = 1,000,000"
each = 100
def setUp(self):
self.nbr_elements = 5000
self.size = 1000000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_p)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.CSR_c = self.A_c.to_csr()
self.CSR_p = self.A_p.to_csr()
self.CSR_s = self.A_s.tocsr()
self.v = np.arange(0, self.size, dtype=np.float64)
class LLMatMatVecBenchmark_2(LLMatMatVecBenchmark):
label = "matvec with 10,000 elements and size = 100,000"
each = 100
def setUp(self):
self.nbr_elements = 10000
self.size = 100000
self.non_contiguity = 10
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size,
self.nbr_elements)
self.CSC_c = self.A_c.to_csc()
self.CSC_s = self.A_s.tocsc()
self.v = np.arange(0,
self.size * self.non_contiguity,
dtype=np.float64)
def A(self, *args, **kwargs):
"""
Evaluate sparse Jacobian of the linear part of the
constraints. Useful to obtain constraint matrix
when problem is a linear programming problem.
"""
vals, rows, cols = super(CySparseAmplModel,
self).A(*args, **kwargs)
A = LLSparseMatrix(nrow=self.ncon,
ncol=self.nvar,
size_hint=vals.size,
store_symmetric=False,
type=types.INT64_T,
dtype=types.FLOAT64_T)
A.put_triplet(rows, cols, vals)
return A
class LLMatFrobeniusNormBenchmark(benchmark.Benchmark):
label = "Frobenius norm with 100 elements and size = 1,000 for a symmetrical matrix"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T,
store_symmetric=True)
construct_sym_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat_sym(self.size, self.size, self.nbr_elements)
construct_sym_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def tearDown(self):
assert self.p_norm == self.c_norm
#assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.p_norm = self.A_p.norm('fro')
return
def test_cysparse(self):
self.c_norm = self.A_c.norm('frob')
return
class LLMatCopyBenchmark(benchmark.Benchmark):
label = "Copy matrix with 100 elements and size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c2[i,j] == self.A_p2[i,j]
def test_pysparse(self):
self.A_p2 = self.A_p.copy()
return
def test_cysparse(self):
self.A_c2 = self.A_c.copy()
return
class LLMatInfiniteNormBenchmark(benchmark.Benchmark):
label = "Infinite norm with 100 elements and size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def tearDown(self):
assert self.p_norm_inf == self.c_norm_inf
#assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.p_norm_inf = self.A_p.norm('inf')
return
def test_cysparse(self):
self.c_norm_inf = self.A_c.norm('inf')
return
class LLMatPutTripletBenchmark(benchmark.Benchmark):
label = "Simple put_triplet with 100 elements, size = 1,000 and put_size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.put_size = 1000
assert self.put_size <= self.size
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.A_sppy = None
self.id1 = np.arange(0, self.put_size, dtype=np.int32)
self.id2 = np.full(self.put_size, 37, dtype=np.int32)
self.b = np.arange(0, self.put_size,dtype=np.float64)
def eachSetUp(self):
self.A_sppy = csarray((self.size, self.size), dtype=np.float64, storagetype='row')
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i, j] == self.A_p[i, j]
def test_pysparse(self):
self.A_p.put(self.b, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.put_triplet(self.id1, self.id2, self.b)
return
def test_sppy(self):
self.A_sppy.put(self.b, self.id1, self.id2, init=True)
def construct_cysparse_matrix(n, nbr_elements):
# int is 32 bits on my machine
A = LLSparseMatrix(size=n, size_hint=nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
for i in xrange(nbr_elements):
A[i % n, (2 * i + 1) % n] = i / 3
return A
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
def setUp(self):
self.nbr_elements = 1000
self.size = 10000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
self.A_s = lil_matrix((self.size, self.size), dtype=np.float64)
self.list_of_matrices = []
self.list_of_matrices.append(self.A_c)
self.list_of_matrices.append(self.A_p)
self.list_of_matrices.append(self.A_s)
construct_random_matrices(self.list_of_matrices, self.size, self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
class LLMatUpdateAddAtBenchmark(benchmark.Benchmark):
label = "Simple update_add_at with 100 elements and size = 1,000 and 100 elements to add"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.nbr_elements_to_add = 100
assert self.nbr_elements_to_add < self.size
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.ones(self.nbr_elements_to_add, dtype=np.int32)
self.id2 = self.v = np.arange(0,
self.nbr_elements_to_add,
dtype=np.int32)
self.val = np.empty(self.nbr_elements_to_add, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.update_add_at(self.val, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.update_add_at(self.id1, self.id2, self.val)
return
def setUp(self):
self.nbr_elements = 100000
self.size = 1000000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
self.A_s = lil_matrix(
(self.size, self.size),
dtype=np.float64) # how do we reserve space in advance?
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.take_size = 1000
assert self.take_size <= self.size
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.arange(0, self.take_size, dtype=np.int32)
self.id2 = np.full(self.take_size, 37, dtype=np.int32)
self.b_c = np.empty((self.take_size, ), dtype=np.float64)
self.b_p = np.empty((self.take_size, ), dtype=np.float64)
def _jac(self, x, lp=False):
"""Helper method to assemble the Jacobian matrix.
See the documentation of :meth:`jac` for more information.
The positional argument `lp` should be set to `True` only if the
problem is known to be a linear program. In this case, the evaluation
of the constraint matrix is cheaper and the argument `x` is ignored.
"""
m = self.m
model = self.model
on = self.original_n
lowerC = np.array(model.lowerC, dtype=np.int64)
nlowerC = model.nlowerC
upperC = np.array(model.upperC, dtype=np.int64)
nupperC = model.nupperC
rangeC = np.array(model.rangeC, dtype=np.int64)
nrangeC = model.nrangeC
# Initialize sparse Jacobian
nnzJ = self.model.nnzj + m
J = LLSparseMatrix(nrow=self.ncon,
ncol=self.nvar,
size_hint=nnzJ,
store_symmetric=False,
itype=types.INT64_T,
dtype=types.FLOAT64_T)
# Insert contribution of general constraints
if lp:
J[:on, :on] = self.model.A()
else:
J[:on, :on] = self.model.jac(x[:on])
# Create a few index lists
rlowerC = np.array(range(nlowerC), dtype=np.int64)
rupperC = np.array(range(nupperC), dtype=np.int64)
rrangeC = np.array(range(nrangeC), dtype=np.int64)
# Insert contribution of slacks on general constraints
J.put_triplet(lowerC, on + rlowerC,
-1.0 * np.ones(nlowerC, dtype=np.float64))
J.put_triplet(upperC, on + nlowerC + rupperC,
-1.0 * np.ones(nupperC, dtype=np.float64))
J.put_triplet(rangeC, on + nlowerC + nupperC + rrangeC,
-1.0 * np.ones(nrangeC, dtype=np.float64))
return J
def setUp(self):
self.nbr_elements = 80000
self.size = 100000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T,
store_symmetric=True)
construct_sym_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat_sym(self.size, self.size, self.nbr_elements)
construct_sym_sparse_matrix(self.A_p, self.size, self.nbr_elements)
class LLMatTakeTripletBenchmark(benchmark.Benchmark):
label = "Simple take_triplet with 100 elements, size = 1,000 and take_size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.take_size = 1000
assert self.take_size <= self.size
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.arange(0, self.take_size, dtype=np.int32)
self.id2 = np.full(self.take_size, 37, dtype=np.int32)
self.b_c = np.empty((self.take_size, ), dtype=np.float64)
self.b_p = np.empty((self.take_size, ), dtype=np.float64)
def tearDown(self):
for i in xrange(self.take_size):
assert self.b_c[i] == self.b_p[i]
def test_pysparse(self):
self.A_p.take(self.b_p, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.take_triplet(self.id1, self.id2, self.b_c)
return
class LLMatTakeTripletBenchmark(benchmark.Benchmark):
label = "Simple take_triplet with 100 elements, size = 1,000 and take_size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.take_size = 1000
assert self.take_size <= self.size
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.arange(0, self.take_size, dtype=np.int32)
self.id2 = np.full(self.take_size, 37, dtype=np.int32)
self.b_c = np.empty((self.take_size,),dtype=np.float64)
self.b_p = np.empty((self.take_size,),dtype=np.float64)
def tearDown(self):
for i in xrange(self.take_size):
assert self.b_c[i] == self.b_p[i]
def test_pysparse(self):
self.A_p.take(self.b_p, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.take_triplet(self.id1, self.id2, self.b_c)
return
class LLMatUpdateAddAtBenchmark(benchmark.Benchmark):
label = "Simple update_add_at with 100 elements and size = 1,000 and 100 elements to add"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.nbr_elements_to_add = 100
assert self.nbr_elements_to_add < self.size
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.ones(self.nbr_elements_to_add, dtype=np.int32)
self.id2 = self.v = np.arange(0, self.nbr_elements_to_add, dtype=np.int32)
self.val = np.empty(self.nbr_elements_to_add, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.update_add_at(self.val, self.id1, self.id2)
return
def test_cysparse(self):
self.A_c.update_add_at(self.id1, self.id2, self.val)
return
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.nbr_elements_to_add = 100
assert self.nbr_elements_to_add < self.size
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
itype=INT32_T,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.ones(self.nbr_elements_to_add, dtype=np.int32)
self.id2 = self.v = np.arange(0,
self.nbr_elements_to_add,
dtype=np.int32)
self.val = np.empty(self.nbr_elements_to_add, dtype=np.float64)
class LLMatScaleBenchmark(benchmark.Benchmark):
label = "Simple scale with 100 elements and size = 1,000 (sigma = 10.47)"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.sigma = 10.47
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c[i,j] == self.A_p[i,j]
def test_pysparse(self):
self.A_p.scale(self.sigma)
return
def test_cysparse(self):
self.A_c.scale(self.sigma)
return
def test_cysparse2(self):
self.A_c *= self.sigma
return
def hess(self, x, z=None, *args, **kwargs):
"""Evaluate Lagrangian Hessian at (x, z)."""
model = self.model
if isinstance(model, QuasiNewtonModel):
return self.hop(x, z, *args, **kwargs)
if z is None:
z = np.zeros(self.m)
on = model.n
H = LLSparseMatrix(size=self.nvar,
size_hint=self.model.nnzh,
store_symmetric=True,
itype=types.INT64_T,
dtype=types.FLOAT64_T)
H[:on, :on] = self.model.hess(x[:on], z, *args, **kwargs)
return H
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.nbr_elements_to_add = 100
assert self.nbr_elements_to_add < self.size
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.ones(self.nbr_elements_to_add, dtype=np.int32)
self.id2 = self.v = np.arange(0, self.nbr_elements_to_add, dtype=np.int32)
self.val = np.empty(self.nbr_elements_to_add, dtype=np.float64)
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.take_size = 1000
assert self.take_size <= self.size
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, itype=INT32_T, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.id1 = np.arange(0, self.take_size, dtype=np.int32)
self.id2 = np.full(self.take_size, 37, dtype=np.int32)
self.b_c = np.empty((self.take_size,),dtype=np.float64)
self.b_p = np.empty((self.take_size,),dtype=np.float64)
class LLMatShiftBenchmark_2(LLMatShiftBenchmark):
label = "Simple shift with 10,000 elements and size = 100,000 (sigma = 10.47)"
each = 100
def setUp(self):
self.nbr_elements = 10000
self.size = 100000
self.sigma = 10.47
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_c2 = self.A_c.copy()
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.A_p2 = self.A_p.copy()
class LLMatFindBenchmark(benchmark.Benchmark):
label = "Simple find with 100 elements, size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# assert self.A_c.nnz == self.A_p.nnz
# # reconstruct initial matrices
# self.A_c_bis = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
# self.A_c_bis.put_triplet(self.A_c_rows, self.A_c_cols, self.A_c_vals)
# self.A_p_bis = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
# self.A_p_bis.put(self.A_p_vals, self.A_p_rows, self.A_p_cols)
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c_bis[i, j] == self.A_p_bis[i, j]
def test_pysparse(self):
self.A_p_vals, self.A_p_rows, self.A_p_cols = self.A_p.find()
return
def test_cysparse(self):
self.A_c_rows, self.A_c_cols, self.A_c_vals = self.A_c.find()
return
class LLMatFindBenchmark(benchmark.Benchmark):
label = "Simple find with 100 elements, size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size,
size_hint=self.nbr_elements,
dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
#def tearDown(self):
# assert self.A_c.nnz == self.A_p.nnz
# # reconstruct initial matrices
# self.A_c_bis = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
# self.A_c_bis.put_triplet(self.A_c_rows, self.A_c_cols, self.A_c_vals)
# self.A_p_bis = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
# self.A_p_bis.put(self.A_p_vals, self.A_p_rows, self.A_p_cols)
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.A_c_bis[i, j] == self.A_p_bis[i, j]
def test_pysparse(self):
self.A_p_vals, self.A_p_rows, self.A_p_cols = self.A_p.find()
return
def test_cysparse(self):
self.A_c_rows, self.A_c_cols, self.A_c_vals = self.A_c.find()
return
class LLMatMatVecTranspBenchmark(benchmark.Benchmark):
label = "A^t * b with 100 elements and size = 1,000"
each = 100
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.v = np.arange(0, self.size, dtype=np.float64)
#def tearDown(self):
# for i in xrange(self.size):
# assert self.w_c[i] == self.w_p[i]
# #assert self.w_c[i] == self.w_s[i]
def test_pysparse(self):
self.w_p = np.empty(self.size, dtype=np.float64)
self.A_p.matvec_transp(self.v, self.w_p)
return
def test_cysparse(self):
self.w_c = self.A_c.T * self.v
return
def test_cysparse2(self):
self.w_c = self.A_c.matvec_transp(self.v)
return
class LLMatMatDotBenchmark(benchmark.Benchmark):
label = "Simple matdot (A^t * B) with 100 elements and size = 1,000"
each = 10
def setUp(self):
self.nbr_elements = 100
self.size = 1000
self.A_c = LLSparseMatrix(size=self.size, size_hint=self.nbr_elements, dtype=FLOAT64_T)
construct_sparse_matrix(self.A_c, self.size, self.nbr_elements)
self.A_p = spmatrix.ll_mat(self.size, self.size, self.nbr_elements)
construct_sparse_matrix(self.A_p, self.size, self.nbr_elements)
self.C_c = None
self.C_c_via_T = None
self.C_p = None
#def tearDown(self):
# for i in xrange(self.size):
# for j in xrange(self.size):
# assert self.C_c[i,j] == self.C_p[i,j]
def test_pysparse(self):
self.C_p = spmatrix.dot(self.A_p, self.A_p)
return
def test_cysparse(self):
self.C_c = self.A_c.matdot_transp(self.A_c)
return
def test_cysparse2(self):
self.C_c = self.A_c.T * self.A_c
return