def lstsq_matrix_test(self):
"""Test least square solution using a matrix with zero column"""
# Initialize input matrices
volumes = [
[24], [20], [20], [20], [21], [30],
[40], [20], [20], [20], [19], [35]
]
promoted = [
[1, 0, 0, 0, 1],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 1, 0, 0],
[1, 0, 0, 1, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 1, 0, 0]
]
volMatrix = Matrix(1, 12)
volMatrix.initialize(volumes)
proMatrix = Matrix(5, 12)
proMatrix.initialize(promoted)
# expected result calculated with scipy.lstsq()
exRes = [
[2.00000000e+01],
[2.18197426e-16],
[1.25000000e+01],
[2.00000000e+01],
[4.00000000e+00]
]
# Calculate least square solution
res = LinearRegression.lstsq(proMatrix, volMatrix)
# Compare if the values of the result are almost equal with the expected ones.
for row in xrange(len(exRes)):
for col in xrange(len(exRes[0])):
self.assertAlmostEqual(exRes[row][col], res.get_value(col, row), self.precision)
def lstsq_test(self):
"""Test the least square method"""
# Initialize input matrices
volumes = [
[24], [20], [20], [20], [21], [30],
[40], [20], [20], [20], [19], [35]
]
promoted = [
[1, 0, 0, 0, 1],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 1, 0, 0, 0],
[1, 0, 1, 0, 0],
[1, 0, 0, 1, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 0, 0, 0, 0],
[1, 1, 0, 0, 0],
[1, 0, 1, 0, 0]
]
expRes = [
[19.999999999999993],
[8.8817841970012523e-15],
[12.500000000000011],
[20.000000000000007],
[4.0]
]
volMatrix = Matrix(1, 12)
volMatrix.initialize(volumes)
proMatrix = Matrix(5, 12)
proMatrix.initialize(promoted)
# Execute least square method and compare result
resMatrix = LinearRegression.lstsq(proMatrix, volMatrix)
for row in range(resMatrix.get_height()):
for col in range(resMatrix.get_width()):
self.assertAlmostEqual(resMatrix.get_value(col, row), expRes[row][col], self.precision)
def lstsq_test(self):
"""Test the least square method"""
# Initialize input matrices
volumes = [[24], [20], [20], [20], [21], [30], [40], [20], [20], [20],
[19], [35]]
promoted = [[1, 0, 0, 0, 1], [1, 0, 0, 0, 0], [1, 0, 0, 0, 0],
[1, 0, 0, 0, 0], [1, 1, 0, 0, 0], [1, 0, 1, 0, 0],
[1, 0, 0, 1, 0], [1, 0, 0, 0, 0], [1, 0, 0, 0, 0],
[1, 0, 0, 0, 0], [1, 1, 0, 0, 0], [1, 0, 1, 0, 0]]
expRes = [[19.999999999999993], [8.8817841970012523e-15],
[12.500000000000011], [20.000000000000007], [4.0]]
volMatrix = Matrix(1, 12)
volMatrix.initialize(volumes)
proMatrix = Matrix(5, 12)
proMatrix.initialize(promoted)
# Execute least square method and compare result
resMatrix = LinearRegression.lstsq(proMatrix, volMatrix)
for row in range(resMatrix.get_height()):
for col in range(resMatrix.get_width()):
self.assertAlmostEqual(resMatrix.get_value(col, row),
expRes[row][col], self.precision)
def lstsq_matrix_test(self):
"""Test least square solution using a matrix with zero column"""
# Initialize input matrices
volumes = [[24], [20], [20], [20], [21], [30], [40], [20], [20], [20],
[19], [35]]
promoted = [[1, 0, 0, 0, 1], [1, 0, 0, 0, 0], [1, 0, 0, 0, 0],
[1, 0, 0, 0, 0], [1, 0, 0, 0, 0], [1, 0, 1, 0, 0],
[1, 0, 0, 1, 0], [1, 0, 0, 0, 0], [1, 0, 0, 0, 0],
[1, 0, 0, 0, 0], [1, 0, 0, 0, 0], [1, 0, 1, 0, 0]]
volMatrix = Matrix(1, 12)
volMatrix.initialize(volumes)
proMatrix = Matrix(5, 12)
proMatrix.initialize(promoted)
# expected result calculated with scipy.lstsq()
exRes = [[2.00000000e+01], [2.18197426e-16], [1.25000000e+01],
[2.00000000e+01], [4.00000000e+00]]
# Calculate least square solution
res = LinearRegression.lstsq(proMatrix, volMatrix)
# Compare if the values of the result are almost equal with the expected ones.
for row in xrange(len(exRes)):
for col in xrange(len(exRes[0])):
self.assertAlmostEqual(exRes[row][col],
res.get_value(col, row), self.precision)