def _visualize_DBSCAN(self, cluster_name):
n_clusters_ = len(set(self.df[cluster_name].tolist()))
if self.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.df,
hue=cluster_name,
palette=DataVisualization.create_categorical_palette(
n_clusters_))
def cluster_comparison_optics(self, radius, neighbors, visualize=False):
self.logger.info("Starting default optics Algorithm on All Points")
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_optics(radius, neighbors)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_comparison_mean_shift(self, bandwidth=None, visualize=False):
self.logger.info("Starting default mean shift Algorithm on All Points")
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_mean_shift(bandwidth)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_comparison_clique(self, intervals, threshold, visualize=False):
self.logger.info("Starting default clique Algorithm on All Points")
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_clique(intervals, threshold)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_comparison_DBSCAN(self, eps, min_samples, visualize=False):
self.logger.info("Starting default DBSCAN Algorithm on All Points")
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_DBSCAN(eps, min_samples)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_comparison_k_means(self, k, visualize=False):
self.logger.info('Starting default k-Means Algorithm on All Points')
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_k_means(k)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_scubi_DBSCAN(self, eps, min_samples, visualize=False):
self.logger.info('Starting SCUBI-DBSCAN Algorithm on Border Points')
border_point_clustering = BorderPointClustering(self.data_frame)
border_point_clustering.cluster_SCUBI_DBSCAN(eps, min_samples,
visualize)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=border_point_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
border_point_clustering.cluster_count))
def cluster_comparison_hierarchical_single_link(self,
cluster_count,
visualize=False):
self.logger.info(
"Starting default hierarchical single link Algorithm on All Points"
)
comparison_clustering = ComparisonClustering(self.data_frame)
comparison_clustering.cluster_hierarchical(cluster_count)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=comparison_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
comparison_clustering.cluster_count))
def cluster_hierarchical_single_link(self,
cluster_count,
modifier=1,
visualize=False):
self.logger.info('Starting Hierarchical Clustering using Single-Link')
border_point_clustering = BorderPointClustering(self.data_frame)
border_point_clustering.cluster_hierarchical(cluster_count, modifier)
if visualize and self.data_frame.dimensions == 2:
DataVisualization.visualize_plot_2d(
self.data_frame.df,
hue=border_point_clustering.clustering_name,
palette=DataVisualization.create_categorical_palette(
border_point_clustering.cluster_count))
return border_point_clustering.clustering_result
def visualize(df,
dimensions,
visualize,
hue,
title=None,
k=None,
show_directions=False):
"""
Common visualization function for scatter plot to show points with exactly ONE certain hue
:param df: The data points
:param dimensions: Dimensions of the data
:param visualize: If the df intsys to be visualized
:param hue: The column to be used as hue from the df frame
:param title: The title of the plot
:return: Displays the plot
"""
if visualize:
if dimensions == 2:
if show_directions:
edges = 0
DataVisualization.visualize_plot_2d(df,
k=k,
hue=hue,
reverse=False,
title=title,
edges=edges)
else:
DataVisualization.visualize_plot_2d(df,
k=k,
hue=hue,
reverse=False,
title=title)
elif dimensions == 3:
DataVisualization.visualize_plot_3d(df, hue=hue)
else:
logging.warning("Cannot visualize " + str(dimensions) +
" dimensions in a scatter plot")