#!/usr/bin/env python
# coding: utf-8

# In[1]:


get_ipython().run_line_magic('matplotlib', 'inline')
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import sklearn


# #K-means Clustering

# In[2]:


import scipy.io


# In[3]:


data1 = scipy.io.loadmat('ex7data2.mat')


# In[4]:


plt.figure(figsize=(8, 6))
plt.scatter(data1['X'][:, 0], data1['X'][:, 1])
plt.show()


# In[5]:


from sklearn.cluster import KMeans


# In[6]:


clf = KMeans(n_clusters=3)
clf.fit(data1['X'])

clf.cluster_centers_


# In[7]:


plt.figure(figsize=(8, 6))
plt.scatter(data1['X'][:, 0], data1['X'][:, 1], c=clf.labels_, alpha=0.3);
plt.scatter(clf.cluster_centers_[:, 0], clf.cluster_centers_[:, 1], c=[1, 2, 3], marker='x', s=300)
plt.show()


# #Image compression with K-means

# In[8]:


import matplotlib.image as mpimg


# In[9]:


bird = mpimg.imread('bird_small.png')


# In[10]:


plt.imshow(bird)
plt.show()


# In[11]:


w, h, d = bird.shape
data = bird.reshape(w * h, d)


# In[12]:


clf2 = KMeans(n_clusters=16)
clf2.fit(data)

print clf2.cluster_centers_


# In[13]:


compressed = clf2.cluster_centers_[clf2.labels_].reshape(w, h, d)


# In[14]:


plt.imshow(compressed)
plt.show()


# #Principal Component Analysis

# In[15]:


data2 = scipy.io.loadmat('ex7data1.mat')


# In[16]:


plt.figure(figsize=(8, 6))
plt.title('Example Dataset 1')
plt.scatter(data2['X'][:, 0], data2['X'][:, 1])
plt.show()


# In[17]:


from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import Pipeline


# In[18]:


clf3 = Pipeline([('zscore', StandardScaler()),
                ('pca', PCA(n_components=1))])


# In[19]:


X = clf3.fit_transform(data2['X'])
print clf3.named_steps['pca'].components_


# In[20]:


X_inverse = clf3.named_steps['zscore'].inverse_transform(clf3.named_steps['pca'].components_ * X)


# In[21]:


plt.figure(figsize=(8, 6))
plt.title('Example Dataset 1')
plt.scatter(data2['X'][:, 0], data2['X'][:, 1])
plt.scatter(X_inverse[:, 0], X_inverse[:, 1], c='r', marker='o')
plt.show()


# In[ ]: