def test_apply_rowwise(self):
A = elem.DistMatrix_d_VR_STAR()
elem.Uniform(A, 100, self.n)
S = sketch.FJLT(self.n, self.sn)
SA = np.zeros((100, self.sn), order='F')
S.apply(A, SA, "rowwise")
def test_apply_colwise(self):
A = elem.DistMatrix_d_VR_STAR()
elem.Uniform(A, self.n, 100)
S = sketch.FJLT(self.n, self.sn)
SA = np.zeros((self.sn, 100), order='F')
S.apply(A, SA, "columnwise")
def test_apply_colwise(self):
A = elem.DistMatrix_d_VR_STAR()
#FIXME: Christos, use your matrix problem factory here
elem.Uniform(A, _M, _N)
#FIXME: A.Matrix will not work in parallel
self.sv = np.linalg.svd(A.Matrix, full_matrices=1, compute_uv=0)
for sketch in self.sketches:
results = test_helper(A, _M, _N, _R, sketch, [self.svd_bound], MPI)
self.check_result(results, str(sketch))
import elem
from skylark import sketch, elemhelper
from mpi4py import MPI
import numpy as np
import time
# Configuration
m = 20000
n = 300
t = 1000
#sketches = { "JLT" : sketch.JLT, "FJLT" : sketch.FJLT, "CWT" : sketch.CWT }
sketches = {"JLT": sketch.JLT, "CWT": sketch.CWT}
# Set up the random regression problem.
A = elem.DistMatrix_d_VR_STAR()
elem.Uniform(A, m, n)
b = elem.DistMatrix_d_VR_STAR()
elem.Uniform(b, m, 1)
# Solve using Elemental
# Elemental currently does not support LS on VR,STAR.
# So we copy.
A1 = elem.DistMatrix_d()
elem.Copy(A, A1)
b1 = elem.DistMatrix_d()
elem.Copy(b, b1)
x = elem.DistMatrix_d(n, 1)
t0 = time.time()
elem.LeastSquares(elem.NORMAL, A1, b1, x)
telp = time.time() - t0