async def do_cluster():
app_state.generator = ClusterGenerator(dsm_mat=app_state.dsm)
c_orig = app_state.generator.cluster(app_state.dsm)
total_coord_cost = app_state.generator.total_coord_cost
cost_history = app_state.generator.cost_history
c = ClusterMatrix.reorder(c_orig)
new_dsm = DSMMatrix.reorder_by_cluster(app_state.dsm, c)
d_new_g = DSMMatrix.place_diag(new_dsm)
app_state.cluster_result = c
app_state.dsm_result = d_new_g
app_state.dsm_annotated = DSMMatrix.annotate_clusters(d_new_g, c)
return {
"dsm": app_state.dsm_result.tojson(),
"labels": app_state.dsm_result.labels,
"cluster": app_state.cluster_result.tojson(),
"dsm_a": app_state.dsm_annotated.tojson(),
}
def test_reorder2(self):
d_mat_2 = np.array([[1, 0, 0, 1], [0, 0, 0, 0], [1, 0, 1, 0],
[0, 0, 1, 1]])
d_list_2 = ["1", "2", "3", "4"]
d_2 = DSMMatrix(d_mat_2, d_list_2)
# print("Original DSM: ")
# print(d_2.mat)
c_mat_2 = np.array([[0, 0, 0, 1], [0, 0, 1, 0], [0, 1, 0, 0],
[1, 0, 0, 0]])
c_2 = ClusterMatrix.from_mat(c_mat_2)
# print("Original Cluster: ")
# print(c_2.mat)
new_d_mat_2 = DSMMatrix.reorder_by_cluster(d_2, c_2)
# print("Reordered DSM: ")
# print(new_d_mat_2.mat)
# print(new_d_mat_2.labels)
val_comp_2 = new_d_mat_2.mat == np.array([[0, 1, 0, 0], [0, 0, 0, 1],
[0, 0, 0, 0], [1, 0, 0, 0]])
lab_comp_2 = new_d_mat_2.labels == ['4', '3', '2', '1']
assert val_comp_2.all()
assert lab_comp_2
def test_reorder1(self):
# d_mat = np.identity(3)
d_mat = np.array([[1, 0, 0, 0], [0, 0, 0, 0], [1, 0, 1, 0],
[0, 0, 1, 1]])
d_list = ["1", "2", "3", "4"]
d = DSMMatrix(d_mat, d_list)
# print("Original DSM: ")
# print(d.mat)
c_mat = np.array([[1, 0, 1, 1], [0, 1, 0, 0], [0, 0, 0, 0],
[0, 0, 0, 0]])
c = ClusterMatrix.from_mat(c_mat)
# print("Original Cluster: ")
# print(c.mat)
new_d_mat = DSMMatrix.reorder_by_cluster(d, c)
# print("Reordered DSM: ")
# print(new_d_mat.mat)
# print(new_d_mat.labels)
val_comp_1 = new_d_mat.mat == np.array([[0, 0, 0, 0], [1, 0, 0, 0],
[0, 1, 0, 0], [0, 0, 0, 0]])
lab_comp_1 = new_d_mat.labels == ['1', '3', '4', '2']
assert val_comp_1.all()
assert lab_comp_1
def _coord_cost(DSM_matrix, cluster_matrix, params):
"""
Calculate the coordination cost of the cluster matrix.
This routine checks all DSM interactions and adds to a total the cost of all intra-cluster interactions. Interactions outside of clusters are assigned a higher cost.
As per Thebeau (2000) the cost formula is, for each interdependent/coupled activity i and j:
(In cluster) (DSM[i] + DSM[j])*cluster_size[cluster_n]^pow_cc
(Out of cluster) (DSM[i] + DSM[j])*DSM_size^pow_cc
Sum each term to obtain the total.
"""
assert isinstance(DSM_matrix, DSMMatrix)
assert isinstance(cluster_matrix, ClusterMatrix)
assert isinstance(params, ClusterParameters)
if params:
pow_cc = params.pow_cc
else:
pow_cc = 1
# get the number of clusters and size of the DSM
n_clusters = cluster_matrix.mat.shape[0]
DSM_size = cluster_matrix.mat.shape[1]
# initialise coordination costs
total_coord_cost = 0
coordination_cost = np.zeros([1, DSM_size])
# dummy variable for reorder function
DSM_labels = [str(x + 1) for x in range(DSM_size)]
# reorder the DSM according to the cluster matrix
new_DSM_matrix = DSMMatrix.reorder_by_cluster(DSM_matrix,
cluster_matrix)
new_DSM_size = new_DSM_matrix.mat.shape[0]
num_cluster_elements = np.sum(cluster_matrix.mat, axis=1, dtype=np.int)
n = 0 # indexing starts from zero
new_cluster_mat = ClusterMatrix.from_mat(
np.zeros([new_DSM_size, new_DSM_size]))
# Create a new cluster matrix that matches the reordered DSM matrix
#
# formerly n_clusters - changed due to shape
# Check again to replicate MATLAB functionality - this is not a permanent fix
for i in range(n_clusters):
if i < new_DSM_size:
new_cluster_mat.mat[i,
n:n + num_cluster_elements[i]] = np.ones(
[1, num_cluster_elements[i]])
n += num_cluster_elements[i]
else:
new_cluster_mat.mat = np.vstack([
new_cluster_mat.mat,
np.zeros(new_cluster_mat.mat.shape[1])
])
new_cluster_mat.mat[i,
n:n + num_cluster_elements[i]] = np.ones(
[1, num_cluster_elements[i]])
n += num_cluster_elements[i]
# import pdb; pdb.set_trace()
# get new cluster size array matching matrix
# new_cluster_size = np.sum(new_cluster_mat.mat, axis=1)
new_cluster_mat.update_cluster_size()
# replace old data with new data for cost calculation
DSM_matrix = new_DSM_matrix
DSM_size = new_DSM_size
cluster_matrix = new_cluster_mat
# cluster_size = new_cluster_size
# get the number of clusters and size of the DSM
n_clusters = cluster_matrix.mat.shape[0]
DSM_size = cluster_matrix.mat.shape[1]
total_coord_cost = 0
coordination_cost = np.zeros([DSM_size, 1])
# Calculate the cost of the solution
for i in range(DSM_size):
for j in range(i + 1, DSM_size):
if DSM_matrix.mat[i, j] > 0 or DSM_matrix.mat[j, i] > 0:
cost_total = 0
for cluster_index in range(n_clusters):
if cluster_matrix.mat[cluster_index,
i] + cluster_matrix.mat[
cluster_index, j] == 2:
# cost_total += (DSM_matrix.mat[i,j] + DSM_matrix.mat[j,i]) * cluster_size[cluster_index]**pow_cc
cost_total += (DSM_matrix.mat[i, j] +
DSM_matrix.mat[j, i]
) * cluster_matrix.cluster_size[
cluster_index]**pow_cc
if cost_total > 0:
cost_c = cost_total
else:
cost_c = (DSM_matrix.mat[i, j] +
DSM_matrix.mat[j, i]) * DSM_size**pow_cc
coordination_cost[i] += cost_c
total_coord_cost = np.sum(coordination_cost)
return total_coord_cost