def test_k3nerror_with_gtm_to_boston(self):
shape_of_map = [10, 10]
shape_of_rbf_centers = [5, 5]
variance_of_rbfs = 4
lambdas = 0.001
iterations = 300
do_display = 0
boston = load_boston()
X = boston.data
#autoscaling
X = (X - X.mean(axis=0)) / X.std(axis=0, ddof=1)
model = gtm(shape_of_map, shape_of_rbf_centers, variance_of_rbfs,
lambdas, iterations, do_display)
model.fit(X)
responsibilities = model.responsibility(X)
means = responsibilities.dot(model.mapgrids)
k = 10
actual = k3nerror(X, means, k)
expected = 13.53674187
self.assertAlmostEqual(expected, actual, places=7)
def get_k3nerror_calculation_speed_with_k_equals(self, k):
shape_of_map = [10, 10]
shape_of_rbf_centers = [5, 5]
variance_of_rbfs = 4
lambdas = 0.001
iterations = 300
do_display = 0
model = gtm(shape_of_map, shape_of_rbf_centers, variance_of_rbfs,
lambdas, iterations, do_display)
model.fit(self.X)
responsibilities = model.responsibility(self.X)
means = responsibilities.dot(model.mapgrids)
def setup_to_iris(self):
shape_of_map = [10, 10]
shape_of_rbf_centers = [5, 5]
variance_of_rbfs = 4
lambdas = 0.001
iterations = 300
do_display = 0
iris = load_iris()
self.X = iris.data
#autoscaling
self.X = (self.X - self.X.mean(axis=0)) / self.X.std(axis=0, ddof=1)
model = gtm(shape_of_map, shape_of_rbf_centers, variance_of_rbfs,
lambdas, iterations, do_display)
model.fit(self.X)
responsibilities = model.responsibility(self.X)
self.means = responsibilities.dot(model.mapgrids)
def test_gtm_with_boston(self):
shape_of_map = [10, 10]
shape_of_rbf_centers = [5, 5]
variance_of_rbfs = 4
lambdas = 0.001
iterations = 300
do_display = 0
boston = load_boston()
X = boston.data
#autoscaling
X = (X - X.mean(axis=0)) / X.std(axis=0, ddof=1)
model = gtm(shape_of_map, shape_of_rbf_centers, variance_of_rbfs,
lambdas, iterations, do_display)
model.fit(X)
responsibilities = model.responsibility(X)
means = responsibilities.dot(model.mapgrids)
actual = means[:10, :]
expected = np.array([
[-0.9121181133961582, 0.3270687429716633],
[-0.7623993487555986, -0.03465058258516577],
[-0.9751389665503443, 0.47031838071377396],
[-0.9881881283635213, 0.17674611687624858],
[-0.9834545901596559, 0.2612741266073518],
[-0.9416833225443894, 0.06562281187680431],
[-0.5371087450312528, -0.8834321955958397],
[-0.39080236753376996, -0.9576518330286727],
[-0.32352324436445645, -0.9372201378955927],
[-0.4773446121858353, -0.9178463400391587],
])
np.testing.assert_allclose(expected, actual, rtol=1e-7, atol=0)
k = 10
actual = k3nerror(X, means, k)
expected = 13.53674187
self.assertAlmostEqual(expected, actual, places=7)
def test_gtm_with_iris(self):
shape_of_map = [10, 10]
shape_of_rbf_centers = [5, 5]
variance_of_rbfs = 4
lambdas = 0.001
iterations = 300
do_display = 0
iris = load_iris()
X = iris.data
#autoscaling
X = (X - X.mean(axis=0)) / X.std(axis=0, ddof=1)
model = gtm(shape_of_map, shape_of_rbf_centers, variance_of_rbfs,
lambdas, iterations, do_display)
model.fit(X)
responsibilities = model.responsibility(X)
means = responsibilities.dot(model.mapgrids)
actual = means[:10, :]
expected = np.array([[-0.6705465933468514, 0.3829717245213705],
[-0.48316715705696234, 0.48573876891327505],
[-0.9326499402098388, -0.4653586650861626],
[-0.8796540759427155, -0.432748196248535],
[-0.808484188863981, 0.1127914668082637],
[-0.8562671930913747, 0.5054770179076009],
[-0.8827856441300941, -0.1900002396932689],
[-0.6567763264305273, 0.43100475126529186],
[-0.9675496106656379, -0.7475745364276133],
[-0.6369141396343855, 0.15331616956496824]])
np.testing.assert_allclose(expected, actual, rtol=1e-7, atol=0)
k = 10
actual = k3nerror(X, means, k)
expected = 0.8332434959
self.assertAlmostEqual(expected, actual, places=7)
# grid search
gridparametersandk3nerror = []
allcalcnumber = len(candidatesofshapeofmap) * len(
candidatesofshapeofrbfcenters) * len(candidatesofvarianceofrbfs) * len(
candidatesoflambdainemalgorithm)
calcnumber = 0
for shapeofmapgrid in candidatesofshapeofmap:
for shapeofrbfcentersgrid in candidatesofshapeofrbfcenters:
for varianceofrbfsgrid in candidatesofvarianceofrbfs:
for lambdainemalgorithmgrid in candidatesoflambdainemalgorithm:
calcnumber = calcnumber + 1
print([calcnumber, allcalcnumber])
# construct GTM model
model = gtm([shapeofmapgrid, shapeofmapgrid],
[shapeofrbfcentersgrid, shapeofrbfcentersgrid],
varianceofrbfsgrid, lambdainemalgorithmgrid,
numberofiterations, desplayflag)
model.fit(inputdataset)
if model.successflag:
# calculate of responsibilities
responsibilities = model.responsibility(inputdataset)
# calculate the mean of responsibilities
means = responsibilities.dot(model.mapgrids)
# calculate k3nerror
k3nerrorofgtm = k3nerror(inputdataset, means, k)
else:
k3nerrorofgtm = 10**100
gridparametersandk3nerror.append([
shapeofmapgrid, shapeofrbfcentersgrid, varianceofrbfsgrid,
lambdainemalgorithmgrid, k3nerrorofgtm
])
parameters_and_k3nerror = []
all_calculation_numbers = len(candidates_of_shape_of_map) * len(
candidates_of_shape_of_rbf_centers) * len(
candidates_of_variance_of_rbfs) * len(
candidates_of_lambda_in_em_algorithm)
calculation_number = 0
for shape_of_map_grid in candidates_of_shape_of_map:
for shape_of_rbf_centers_grid in candidates_of_shape_of_rbf_centers:
for variance_of_rbfs_grid in candidates_of_variance_of_rbfs:
for lambda_in_em_algorithm_grid in candidates_of_lambda_in_em_algorithm:
calculation_number += 1
print([calculation_number, all_calculation_numbers])
# construct GTM model
model = gtm(
[shape_of_map_grid, shape_of_map_grid],
[shape_of_rbf_centers_grid, shape_of_rbf_centers_grid],
variance_of_rbfs_grid, lambda_in_em_algorithm_grid,
number_of_iterations, display_flag)
model.fit(input_dataset)
if model.success_flag:
# calculate of responsibilities
responsibilities = model.responsibility(input_dataset)
# calculate the mean of responsibilities
means = responsibilities.dot(model.map_grids)
# calculate k3n-error
k3nerror_of_gtm = k3nerror(input_dataset, means,
k_in_k3nerror)
else:
k3nerror_of_gtm = 10**100
parameters_and_k3nerror.append([
shape_of_map_grid, shape_of_rbf_centers_grid,