def get_cluster_graph(self, df_result, label):
gv = GraphVisualizer()
gv.set_plotly()
data_meta_list = []
for i in OrderedDict.fromkeys(df_result[label]):
content_label_list = []
for content_label in df_result[df_result.predict == i]["content"]:
if len(content_label) > 30:
content_label = content_label[:30] + "..."
content_label_list.append(content_label)
else:
content_label_list.append(content_label)
data_meta = {
"data_name": i,
"x_data": df_result[df_result[label] == i]["x"],
"y_data": df_result[df_result[label] == i]["y"],
"label": content_label_list
}
data_meta_list.append(data_meta)
graph_meta = {
"title": "Cluter Graph - " + label,
"x_name": "TSNE X",
"y_name": "TSNE Y"
}
return gv.draw_scatter(data_meta_list, graph_meta)
def get_kmeans_graph(self, df_result, label):
gv = GraphVisualizer()
gv.set_plotly()
data_meta_list = []
for predict in list(OrderedDict.fromkeys(df_result[label])):
data_meta = {
"data_name": predict,
"x_data": df_result[df_result[label]==predict]["x"],
"y_data": df_result[df_result[label]==predict]["y"],
"label": predict
}
data_meta_list.append(data_meta)
graph_meta = {
"title": "K-Means Clutering Graph - " + label,
"x_name": "TSNE X",
"y_name": "TSNE Y"
}
return gv.draw_scatter(data_meta_list, graph_meta)