Point Cloud Weights
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
# Dimension reduction and clustering libraries
import umap
import sklearn.cluster as cluster
data = np.genfromtxt("ProjectionWs.txt", delimiter=",")
label = np.genfromtxt("train_c.txt", delimiter=",")
label
# label_train = sio.loadmat('Open3Dpointnet/train_label.mat')
# label = label_train['Classification_labels'][0]
standard_embedding = umap.UMAP(random_state=42).fit_transform(data)
plt.scatter(standard_embedding[:, 0], standard_embedding[:, 1], c=label, s=0.1, cmap='Spectral');
plt.legend()
plt.xlabel('PC1')
plt.ylabel('PC2')
import umap.plot
mapper = umap.UMAP().fit(data)
umap.plot.points(mapper, labels=label, color_key_cmap='Paired', background='black')
kmeans_labels = cluster.KMeans(n_clusters=10).fit_predict(data)
plt.scatter(standard_embedding[:, 0], standard_embedding[:, 1], c=kmeans_labels, s=0.5, cmap='Spectral');
plt.legend()
