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 graph_matrix(gen_matrix, cluster_matrix=None,
x_title='Element', y_title='Element', graph_title='DSM Matrix',
x_tcklabel=None, y_tcklabel=None, print_flag=0):
""" Graph the DSM and cluster matrix.
"""
assert isinstance(gen_matrix, DSMMatrix) or isinstance(gen_matrix, ClusterMatrix)
# if not isinstance(cluster_matrix, ClusterMatrix)
ds_mat = gen_matrix.mat
(row_in, col_out) = np.where(ds_mat)
m_value = ds_mat[ds_mat != 0]
max_row = np.max(row_in)
max_col = np.max(col_out)
axis = np.array([0, max_col, max_row]);
max_value = np.max(m_value)
data_scale = np.ceil(500/max_value)/20
if x_tcklabel is None and isinstance(gen_matrix, DSMMatrix):
x_tcklabel = gen_matrix.labels
elif isinstance(x_tcklabel, list):
if len(x_tcklabel) != ds_mat.shape[0]:
x_tcklabel = [str(x+1) for x in range(ds_mat.shape[0])]
else:
x_tcklabel = [str(x+1) for x in range(ds_mat.shape[0])]
if y_tcklabel is None and isinstance(gen_matrix, DSMMatrix):
y_tcklabel = gen_matrix.labels
elif isinstance(y_tcklabel, list):
if len(y_tcklabel) != ds_mat.shape[0]:
y_tcklabel = [str(x+1) for x in range(ds_mat.shape[0])]
else:
y_tcklabel = [str(x+1) for x in range(ds_mat.shape[0])]
if isinstance(cluster_matrix, ClusterMatrix):
new_gen_matrix = DSMMatrix.annotate_clusters(gen_matrix, cluster_matrix)
orig_ds_mat = gen_matrix.mat
ds_mat = new_gen_matrix.mat
fig, ax = plt.subplots()
im = ax.imshow(ds_mat, cmap=plt.cm.Blues)
ax.set_xticks(np.arange(len(x_tcklabel)))
ax.set_yticks(np.arange(len(y_tcklabel)))
ax.set_xticklabels(x_tcklabel)
ax.set_yticklabels(y_tcklabel)
ax.set_xlabel(x_title)
ax.set_ylabel(y_title)
ax.xaxis.set_label_position('top')
ax.tick_params(top=True, bottom=False,
labeltop=True, labelbottom=False)
# plt.setp(ax.get_xticklabels(), rotation=90, ha="right",
# rotation_mode="anchor")
ax.set_title(graph_title)
fig.tight_layout()
plt.draw()