def calculate_analysis_values(embeddings_linkage, true_clusters):
"""
Calculates the analysis values out of the embedding linkage.
:param embeddings_linkage: The linkage we calculate the values for.
:param true_clusters: The validation clusters
:return: misclassification rate, homogeneity Score, completeness score and the thresholds.
"""
logger = get_logger('analysis', logging.INFO)
logger.info('Calculate scores')
thresholds = embeddings_linkage[:, 2]
threshold_shape = thresholds.shape
# Initialize output
mrs = np.ones(threshold_shape)
homogeneity_scores = np.ones(threshold_shape)
completeness_scores = np.ones(threshold_shape)
# Loop over all possible clustering
for i, threshold in enumerate(thresholds):
predicted_clusters = fcluster(embeddings_linkage, threshold, 'distance')
# Calculate different analysis's
mrs[i] = misclassification_rate(true_clusters, predicted_clusters)
homogeneity_scores[i] = homogeneity_score(true_clusters, predicted_clusters)
completeness_scores[i] = completeness_score(true_clusters, predicted_clusters)
return mrs, homogeneity_scores, completeness_scores, thresholds
def calculate_analysis_values(predicted_clusters, true_cluster):
"""
Calculates the analysis values out of the predicted_clusters.
:param predicted_clusters: The predicted Clusters of the Network.
:param true_clusters: The validation clusters
:return: misclassification rate, homogeneity Score, completeness score and the thresholds.
"""
logger = get_logger('analysis', logging.INFO)
logger.info('Calculate scores')
# Initialize output
mrs = np.ones(len(true_cluster))
homogeneity_scores = np.ones(len(true_cluster))
completeness_scores = np.ones(len(true_cluster))
# Loop over all possible clustering
for i, predicted_cluster in enumerate(predicted_clusters):
# Calculate different analysis's
mrs[i] = misclassification_rate(true_cluster, predicted_cluster)
homogeneity_scores[i] = homogeneity_score(true_cluster,
predicted_cluster)
completeness_scores[i] = completeness_score(true_cluster,
predicted_cluster)
return mrs, homogeneity_scores, completeness_scores
def calculate_analysis_values(predicted_clusters, true_cluster, cluster_count=None, mr_list=None, mr_dict_index=None):
"""
Calculates the analysis values out of the predicted_clusters.
:param predicted_clusters: The predicted Clusters of the Network.
:param true_clusters: The validation clusters
:return: misclassification rate, homogeneity Score, completeness score and the thresholds.
"""
logger = get_logger('analysis', logging.INFO)
logger.info('Calculate scores')
#
# print("------------------>>>>>>>>>>> before incremental true clusters\n")
# print(true_cluster)
# for i in range(len(true_cluster)):
# true_cluster[i] += 1
#
# print("------------------>>>>>>>>>>> after incremental true clusters\n")
# print(true_cluster)
# Initialize output
mrs = np.ones(len(true_cluster))
homogeneity_scores = np.ones(len(true_cluster))
completeness_scores = np.ones(len(true_cluster))
# Loop over all possible clustering
for i, predicted_cluster in enumerate(predicted_clusters):
# Calculate different analysis's
mrs[i] = misclassification_rate(true_cluster, predicted_cluster)
homogeneity_scores[i] = homogeneity_score(true_cluster, predicted_cluster)
completeness_scores[i] = completeness_score(true_cluster, predicted_cluster)
#print("---------------------------------->>>>>>>>>>>>>>>>>>>>...")
#print(i, predicted_cluster)
if cluster_count is not None and ((max(predicted_cluster) == cluster_count and mr_dict_index == 'AHC') or
(max(predicted_cluster) == cluster_count -1 and mr_dict_index == 'K-MEANS') or
(max(predicted_cluster) == cluster_count -1 and mr_dict_index == 'SP') or
(max(predicted_cluster) == cluster_count -1 and mr_dict_index == 'K-MEDOIDS')):
if mr_dict_index is not None:
mr_list[mr_dict_index].append(mrs[i])
else:
mr_list[cluster_count] = mrs[i]
elif cluster_count is not None and mr_dict_index == "DS":
if mr_dict_index is not None:
mr_list[mr_dict_index].append(mrs[i])
else:
mr_list[cluster_count] = mrs[i]
#print(">>>>>>>>>>>>>>>>>>>>>>>>>> Predicted and true clusters")
#print(predicted_clusters, true_cluster)
return mrs, homogeneity_scores, completeness_scores
def _calculate_analysis_values(predicted_clusters, true_cluster, times):
"""
Calculates the analysis values out of the predicted_clusters.
:param predicted_clusters: The predicted Clusters of the Network.
:param true_clusters: The validation clusters
:return: the results of all metrics as a 2D array where i is the index of the metric and j is the index of a
specific result
"""
logger = get_logger('analysis', logging.INFO)
logger.info('Calculate scores')
# Initialize output
metric_results = [None] * len(metric_names)
for m, min_value in enumerate(metric_min_values):
if min_value == 1:
metric_results[m] = np.ones(len(true_cluster))
else:
metric_results[m] = np.zeros((len(true_cluster)))
# Loop over all possible clustering
for i, predicted_cluster in enumerate(predicted_clusters):
logger.info('Calculated Scores for {}/{} predicted clusters'.format(
i, len(predicted_clusters)))
# Calculate different analysis's
metric_results[0][i] = misclassification_rate(true_cluster,
predicted_cluster)
metric_results[1][i] = average_cluster_purity(true_cluster,
predicted_cluster)
metric_results[2][i] = adjusted_rand_index(true_cluster,
predicted_cluster)
metric_results[3][i] = diarization_error_rate(true_cluster,
predicted_cluster, times)
return metric_results
def calculate_analysis_values(predicted_clusters, true_cluster, alorithm, vector,j , cluster_count=None, mr_list=None , mr_list2=None):
"""
Calculates the analysis values out of the predicted_clusters.
:param predicted_clusters: The predicted Clusters of the Network.
:param true_clusters: The validation clusters
:return: misclassification rate, homogeneity Score, completeness score and the thresholds.
"""
logger = get_logger('analysis', logging.INFO)
logger.info('Calculate scores')
#
# print("------------------>>>>>>>>>>> before incremental true clusters\n")
# print(true_cluster)
# for i in range(len(true_cluster)):
# true_cluster[i] += 1
#
# print("------------------>>>>>>>>>>> after incremental true clusters\n")
# print(true_cluster)
# Initialize output
mrs = np.ones(len(true_cluster))
homogeneity_scores = np.ones(len(true_cluster))
completeness_scores = np.ones(len(true_cluster))
# Loop over all possible clustering
for i, predicted_cluster in enumerate(predicted_clusters):
# Calculate different analysis's
# print(i)
# print("\n")
# print(predicted_cluster)
if alorithm == "K_Means_Clustering":
print("True Clusters")
print(true_cluster)
print("\n")
print("Predicted Clusters")
print(predicted_clusters)
print("\n")
mrs[i] = misclassification_rate(true_cluster, predicted_cluster)
print("...................MR value of K means................................\n")
print(mrs[i])
print("\n")
if cluster_count is not None:
temp = str(cluster_count) + "_" + str(vector) + "_" + str(j)
mr_list2[temp] = mrs[i]
print(str(mrs[i]) + " added to mr_list2\n")
if alorithm == "Agglomerative_Hierachial_Clustering":
mrs[i] = misclassification_rate(true_cluster, predicted_cluster)
print(mrs[i])
if cluster_count is not None and (max(predicted_cluster) == cluster_count):
temp = str(cluster_count) + "_" + str(vector) + "_" + str(j)
mr_list[temp] = mrs[i]
print("...................MR value of Hirechial Clustering................................\n")
print(mrs[i])
print("\n")
print(str(mrs[i]) + " added to mr_list\n")
homogeneity_scores[i] = homogeneity_score(true_cluster, predicted_cluster)
completeness_scores[i] = completeness_score(true_cluster, predicted_cluster)
return mrs, homogeneity_scores, completeness_scores ,mr_list