def get_distance_matrix(self):
"""!
@brief Calculates distance matrix (U-matrix).
@details The U-Matrix visualizes based on the distance in input space between a weight vector and its neighbors on map.
@return (list) Distance matrix (U-matrix).
@see show_distance_matrix()
@see get_density_matrix()
"""
if self.__ccore_som_pointer is not None:
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
if self._conn_type != type_conn.func_neighbor:
self._neighbors = wrapper.som_get_neighbors(self.__ccore_som_pointer)
distance_matrix = [[0.0] * self._cols for i in range(self._rows)]
for i in range(self._rows):
for j in range(self._cols):
neuron_index = i * self._cols + j
if self._conn_type == type_conn.func_neighbor:
self._create_connections(type_conn.grid_eight)
for neighbor_index in self._neighbors[neuron_index]:
distance_matrix[i][j] += euclidean_distance_square(self._weights[neuron_index], self._weights[neighbor_index])
distance_matrix[i][j] /= len(self._neighbors[neuron_index])
return distance_matrix
def get_distance_matrix(self):
"""!
@brief Calculates distance matrix (U-matrix).
@details The U-Matrix visualizes based on the distance in input space between a weight vector and its neighbors on map.
@return (list) Distance matrix (U-matrix).
@see show_distance_matrix()
@see get_density_matrix()
"""
if (self.__ccore_som_pointer is not None):
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
if (self._conn_type != type_conn.func_neighbor):
self._neighbors = wrapper.som_get_neighbors(
self.__ccore_som_pointer)
distance_matrix = [[0.0] * self._cols for i in range(self._rows)]
for i in range(self._rows):
for j in range(self._cols):
neuron_index = i * self._cols + j
if (self._conn_type == type_conn.func_neighbor):
self._create_connections(type_conn.grid_eight)
for neighbor_index in self._neighbors[neuron_index]:
distance_matrix[i][j] += euclidean_distance_sqrt(
self._weights[neuron_index],
self._weights[neighbor_index])
distance_matrix[i][j] /= len(self._neighbors[neuron_index])
return distance_matrix
def weights(self):
"""!
@return (list) Weights of each neuron.
"""
if (self.__ccore_som_pointer is not None):
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
return self._weights
def weights(self):
"""!
@return (list) Weights of each neuron.
"""
if (self.__ccore_som_pointer is not None):
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer);
return self._weights;
def weights(self):
"""!
@brief Return weight of each neuron.
@return (list) Weights of each neuron.
"""
if self.__ccore_som_pointer is not None:
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
return self._weights
def get_density_matrix(self, surface_divider=20.0):
"""!
@brief Calculates density matrix (P-Matrix).
@param[in] surface_divider (double): Divider in each dimension that affect radius for density measurement.
@return (list) Density matrix (P-Matrix).
@see get_distance_matrix()
"""
if (self.__ccore_som_pointer is not None):
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
density_matrix = [[0] * self._cols for i in range(self._rows)]
dimension = len(self._weights[0])
dim_max = [float('-Inf')] * dimension
dim_min = [float('Inf')] * dimension
for weight in self._weights:
for index_dim in range(dimension):
if (weight[index_dim] > dim_max[index_dim]):
dim_max[index_dim] = weight[index_dim]
if (weight[index_dim] < dim_min[index_dim]):
dim_min[index_dim] = weight[index_dim]
radius = [0.0] * len(self._weights[0])
for index_dim in range(dimension):
radius[index_dim] = (dim_max[index_dim] -
dim_min[index_dim]) / surface_divider
for point in self._data:
for index_neuron in range(len(self)):
point_covered = True
for index_dim in range(dimension):
if (abs(point[index_dim] -
self._weights[index_neuron][index_dim]) >
radius[index_dim]):
point_covered = False
break
row = math.floor(index_neuron / self._cols)
col = index_neuron - row * self._cols
if (point_covered is True):
density_matrix[row][col] += 1
return density_matrix
def get_density_matrix(self, surface_divider = 20.0):
"""!
@brief Calculates density matrix (P-Matrix).
@param[in] surface_divider (double): Divider in each dimension that affect radius for density measurement.
@return (list) Density matrix (P-Matrix).
@see get_distance_matrix()
"""
if self.__ccore_som_pointer is not None:
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
density_matrix = [[0] * self._cols for i in range(self._rows)]
dimension = len(self._weights[0])
dim_max = [ float('-Inf') ] * dimension
dim_min = [ float('Inf') ] * dimension
for weight in self._weights:
for index_dim in range(dimension):
if weight[index_dim] > dim_max[index_dim]:
dim_max[index_dim] = weight[index_dim]
if weight[index_dim] < dim_min[index_dim]:
dim_min[index_dim] = weight[index_dim]
radius = [0.0] * len(self._weights[0])
for index_dim in range(dimension):
radius[index_dim] = ( dim_max[index_dim] - dim_min[index_dim] ) / surface_divider
## TODO: do not use data
for point in self._data:
for index_neuron in range(len(self)):
point_covered = True
for index_dim in range(dimension):
if abs(point[index_dim] - self._weights[index_neuron][index_dim]) > radius[index_dim]:
point_covered = False
break
row = int(math.floor(index_neuron / self._cols))
col = index_neuron - row * self._cols
if point_covered is True:
density_matrix[row][col] += 1
return density_matrix
def get_density_matrix(self, surface_divider = 20.0):
"""!
@brief Calculates density matrix (P-Matrix).
@param[in] surface_divider (double): Divider in each dimension that affect radius for density measurement.
@return (list) Density matrix (P-Matrix).
@see get_distance_matrix()
"""
if (self.__ccore_som_pointer is not None):
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer);
density_matrix = [ [0] * self._cols for i in range(self._rows) ];
dimension = len(self._weights[0]);
dim_max = [ float('-Inf') ] * dimension;
dim_min = [ float('Inf') ] * dimension;
for weight in self._weights:
for index_dim in range(dimension):
if (weight[index_dim] > dim_max[index_dim]):
dim_max[index_dim] = weight[index_dim];
if (weight[index_dim] < dim_min[index_dim]):
dim_min[index_dim] = weight[index_dim];
radius = [0.0] * len(self._weights[0]);
for index_dim in range(dimension):
radius[index_dim] = ( dim_max[index_dim] - dim_min[index_dim] ) / surface_divider;
for point in self._data:
for index_neuron in range(len(self)):
point_covered = True;
for index_dim in range(dimension):
if (abs(point[index_dim] - self._weights[index_neuron][index_dim]) > radius[index_dim]):
point_covered = False;
break;
row = math.floor(index_neuron / self._cols);
col = index_neuron - row * self._cols;
if (point_covered is True):
density_matrix[row][col] += 1;
return density_matrix;
def show_network(self,
awards=False,
belongs=False,
coupling=True,
dataset=True,
marker_type='o'):
"""!
@brief Shows neurons in the dimension of data.
@param[in] awards (bool): If True - displays how many objects won each neuron.
@param[in] belongs (bool): If True - marks each won object by according index of neuron-winner (only when dataset is displayed too).
@param[in] coupling (bool): If True - displays connections between neurons (except case when function neighbor is used).
@param[in] dataset (bool): If True - displays inputs data set.
@param[in] marker_type (string): Defines marker that is used for dispaying neurons in the network.
"""
if (self.__ccore_som_pointer is not None):
self._size = wrapper.som_get_size(self.__ccore_som_pointer)
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
self._neighbors = wrapper.som_get_neighbors(
self.__ccore_som_pointer)
self._award = wrapper.som_get_awards(self.__ccore_som_pointer)
dimension = len(self._weights[0])
fig = plt.figure()
axes = None
# Check for dimensions
if ((dimension == 1) or (dimension == 2)):
axes = fig.add_subplot(111)
elif (dimension == 3):
axes = fig.gca(projection='3d')
else:
raise NameError(
'Dwawer supports only 1D, 2D and 3D data representation')
# Show data
if ((self._data is not None) and (dataset is True)):
for x in self._data:
if (dimension == 1):
axes.plot(x[0], 0.0, 'b|', ms=30)
elif (dimension == 2):
axes.plot(x[0], x[1], 'b.')
elif (dimension == 3):
axes.scatter(x[0], x[1], x[2], c='b', marker='.')
# Show neurons
for index in range(self._size):
color = 'g'
if (self._award[index] == 0): color = 'y'
if (dimension == 1):
axes.plot(self._weights[index][0], 0.0, color + marker_type)
if (awards == True):
location = '{0}'.format(self._award[index])
axes.text(self._weights[index][0],
0.0,
location,
color='black',
fontsize=10)
if (belongs == True):
location = '{0}'.format(index)
axes.text(self._weights[index][0],
0.0,
location,
color='black',
fontsize=12)
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]]
axes.text(point[0],
0.0,
location,
color='blue',
fontsize=10)
if (dimension == 2):
axes.plot(self._weights[index][0], self._weights[index][1],
color + marker_type)
if (awards == True):
location = '{0}'.format(self._award[index])
axes.text(self._weights[index][0],
self._weights[index][1],
location,
color='black',
fontsize=10)
if (belongs == True):
location = '{0}'.format(index)
axes.text(self._weights[index][0],
self._weights[index][1],
location,
color='black',
fontsize=12)
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]]
axes.text(point[0],
point[1],
location,
color='blue',
fontsize=10)
if ((self._conn_type != type_conn.func_neighbor)
and (coupling != False)):
for neighbor in self._neighbors[index]:
if (neighbor > index):
axes.plot([
self._weights[index][0],
self._weights[neighbor][0]
], [
self._weights[index][1],
self._weights[neighbor][1]
],
'g',
linewidth=0.5)
elif (dimension == 3):
axes.scatter(self._weights[index][0],
self._weights[index][1],
self._weights[index][2],
c=color,
marker=marker_type)
if ((self._conn_type != type_conn.func_neighbor)
and (coupling != False)):
for neighbor in self._neighbors[index]:
if (neighbor > index):
axes.plot([
self._weights[index][0],
self._weights[neighbor][0]
], [
self._weights[index][1],
self._weights[neighbor][1]
], [
self._weights[index][2],
self._weights[neighbor][2]
],
'g-',
linewidth=0.5)
plt.title("Network Structure")
plt.grid()
plt.show()
def __download_dump_from_ccore(self):
self._location = self.__initialize_locations(self._rows, self._cols)
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
self._award = wrapper.som_get_awards(self.__ccore_som_pointer)
self._capture_objects = wrapper.som_get_capture_objects(self.__ccore_som_pointer)
def __download_dump_from_ccore(self):
self._location = self.__initialize_locations(self._rows, self._cols)
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
self._award = wrapper.som_get_awards(self.__ccore_som_pointer)
self._capture_objects = wrapper.som_get_capture_objects(self.__ccore_som_pointer)
def show_network(self, awards = False, belongs = False, coupling = True, dataset = True, marker_type = 'o'):
"""!
@brief Shows neurons in the dimension of data.
@param[in] awards (bool): If True - displays how many objects won each neuron.
@param[in] belongs (bool): If True - marks each won object by according index of neuron-winner (only when dataset is displayed too).
@param[in] coupling (bool): If True - displays connections between neurons (except case when function neighbor is used).
@param[in] dataset (bool): If True - displays inputs data set.
@param[in] marker_type (string): Defines marker that is used for dispaying neurons in the network.
"""
if self.__ccore_som_pointer is not None:
self._size = wrapper.som_get_size(self.__ccore_som_pointer)
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer)
self._neighbors = wrapper.som_get_neighbors(self.__ccore_som_pointer)
self._award = wrapper.som_get_awards(self.__ccore_som_pointer)
dimension = len(self._weights[0])
fig = plt.figure()
# Check for dimensions
if (dimension == 1) or (dimension == 2):
axes = fig.add_subplot(111)
elif dimension == 3:
axes = fig.gca(projection='3d')
else:
raise NotImplementedError('Impossible to show network in data-space that is differ from 1D, 2D or 3D.')
if (self._data is not None) and (dataset is True):
for x in self._data:
if dimension == 1:
axes.plot(x[0], 0.0, 'b|', ms = 30)
elif dimension == 2:
axes.plot(x[0], x[1], 'b.')
elif dimension == 3:
axes.scatter(x[0], x[1], x[2], c = 'b', marker = '.')
# Show neurons
for index in range(self._size):
color = 'g'
if self._award[index] == 0:
color = 'y'
if dimension == 1:
axes.plot(self._weights[index][0], 0.0, color + marker_type)
if awards:
location = '{0}'.format(self._award[index])
axes.text(self._weights[index][0], 0.0, location, color='black', fontsize = 10)
if belongs and self._data is not None:
location = '{0}'.format(index)
axes.text(self._weights[index][0], 0.0, location, color='black', fontsize = 12)
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]]
axes.text(point[0], 0.0, location, color='blue', fontsize = 10)
if dimension == 2:
axes.plot(self._weights[index][0], self._weights[index][1], color + marker_type)
if awards:
location = '{0}'.format(self._award[index])
axes.text(self._weights[index][0], self._weights[index][1], location, color='black', fontsize=10)
if belongs and self._data is not None:
location = '{0}'.format(index)
axes.text(self._weights[index][0], self._weights[index][1], location, color='black', fontsize=12)
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]]
axes.text(point[0], point[1], location, color='blue', fontsize=10)
if (self._conn_type != type_conn.func_neighbor) and (coupling != False):
for neighbor in self._neighbors[index]:
if neighbor > index:
axes.plot([self._weights[index][0], self._weights[neighbor][0]],
[self._weights[index][1], self._weights[neighbor][1]],
'g', linewidth=0.5)
elif dimension == 3:
axes.scatter(self._weights[index][0], self._weights[index][1], self._weights[index][2], c=color, marker=marker_type)
if (self._conn_type != type_conn.func_neighbor) and (coupling != False):
for neighbor in self._neighbors[index]:
if neighbor > index:
axes.plot([self._weights[index][0], self._weights[neighbor][0]],
[self._weights[index][1], self._weights[neighbor][1]],
[self._weights[index][2], self._weights[neighbor][2]],
'g-', linewidth=0.5)
plt.title("Network Structure")
plt.grid()
plt.show()
def show_network(self, awards = False, belongs = False, coupling = True, dataset = True, marker_type = 'o'):
"""!
@brief Shows neurons in the dimension of data.
@param[in] awards (bool): If True - displays how many objects won each neuron.
@param[in] belongs (bool): If True - marks each won object by according index of neuron-winner (only when dataset is displayed too).
@param[in] coupling (bool): If True - displays connections between neurons (except case when function neighbor is used).
@param[in] dataset (bool): If True - displays inputs data set.
@param[in] marker_type (string): Defines marker that is used for dispaying neurons in the network.
"""
if (self.__ccore_som_pointer is not None):
self._size = wrapper.som_get_size(self.__ccore_som_pointer);
self._weights = wrapper.som_get_weights(self.__ccore_som_pointer);
self._neighbors = wrapper.som_get_neighbors(self.__ccore_som_pointer);
self._award = wrapper.som_get_awards(self.__ccore_som_pointer);
dimension = len(self._weights[0]);
fig = plt.figure();
axes = None;
# Check for dimensions
if ( (dimension == 1) or (dimension == 2) ):
axes = fig.add_subplot(111);
elif (dimension == 3):
axes = fig.gca(projection='3d');
else:
raise NameError('Dwawer supports only 1D, 2D and 3D data representation');
# Show data
if (dataset == True):
for x in self._data:
if (dimension == 1):
axes.plot(x[0], 0.0, 'b|', ms = 30);
elif (dimension == 2):
axes.plot(x[0], x[1], 'b.');
elif (dimension == 3):
axes.scatter(x[0], x[1], x[2], c = 'b', marker = '.');
# Show neurons
for index in range(self._size):
color = 'g';
if (self._award[index] == 0): color = 'y';
if (dimension == 1):
axes.plot(self._weights[index][0], 0.0, color + marker_type);
if (awards == True):
location = '{0}'.format(self._award[index]);
axes.text(self._weights[index][0], 0.0, location, color='black', fontsize = 10);
if (belongs == True):
location = '{0}'.format(index);
axes.text(self._weights[index][0], 0.0, location, color='black', fontsize = 12);
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]];
axes.text(point[0], 0.0, location, color='blue', fontsize = 10);
if (dimension == 2):
axes.plot(self._weights[index][0], self._weights[index][1], color + marker_type);
if (awards == True):
location = '{0}'.format(self._award[index]);
axes.text(self._weights[index][0], self._weights[index][1], location, color='black', fontsize = 10);
if (belongs == True):
location = '{0}'.format(index);
axes.text(self._weights[index][0], self._weights[index][1], location, color='black', fontsize = 12);
for k in range(len(self._capture_objects[index])):
point = self._data[self._capture_objects[index][k]];
axes.text(point[0], point[1], location, color='blue', fontsize = 10);
if ( (self._conn_type != type_conn.func_neighbor) and (coupling != False) ):
for neighbor in self._neighbors[index]:
if (neighbor > index):
axes.plot([self._weights[index][0], self._weights[neighbor][0]], [self._weights[index][1], self._weights[neighbor][1]], 'g', linewidth = 0.5);
elif (dimension == 3):
axes.scatter(self._weights[index][0], self._weights[index][1], self._weights[index][2], c = color, marker = marker_type);
if ( (self._conn_type != type_conn.func_neighbor) and (coupling != False) ):
for neighbor in self._neighbors[index]:
if (neighbor > index):
axes.plot([self._weights[index][0], self._weights[neighbor][0]], [self._weights[index][1], self._weights[neighbor][1]], [self._weights[index][2], self._weights[neighbor][2]], 'g-', linewidth = 0.5);
plt.title("Network Structure");
plt.grid();
plt.show();