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

# # k-평균

# <table align="left">
#   <td>
#     <a target="_blank" href="https://colab.research.google.com/github/rickiepark/hg-mldl/blob/master/6-2.ipynb"><img src="https://www.tensorflow.org/images/colab_logo_32px.png" />구글 코랩에서 실행하기</a>
#   </td>
# </table>

# ## KMeans 클래스

# In[1]:


get_ipython().system('wget https://bit.ly/fruits_300_data -O fruits_300.npy')


# In[2]:


import numpy as np

fruits = np.load('fruits_300.npy')
fruits_2d = fruits.reshape(-1, 100*100)


# In[3]:


from sklearn.cluster import KMeans

km = KMeans(n_clusters=3, random_state=42)
km.fit(fruits_2d)


# In[4]:


print(km.labels_)


# In[5]:


print(np.unique(km.labels_, return_counts=True))


# In[6]:


import matplotlib.pyplot as plt

def draw_fruits(arr, ratio=1):
    n = len(arr)    # n은 샘플 개수입니다
    # 한 줄에 10개씩 이미지를 그립니다. 샘플 개수를 10으로 나누어 전체 행 개수를 계산합니다.
    rows = int(np.ceil(n/10))
    # 행이 1개 이면 열 개수는 샘플 개수입니다. 그렇지 않으면 10개입니다.
    cols = n if rows < 2 else 10
    fig, axs = plt.subplots(rows, cols,
                            figsize=(cols*ratio, rows*ratio), squeeze=False)
    for i in range(rows):
        for j in range(cols):
            if i*10 + j < n:    # n 개까지만 그립니다.
                axs[i, j].imshow(arr[i*10 + j], cmap='gray_r')
            axs[i, j].axis('off')
    plt.show()


# In[7]:


draw_fruits(fruits[km.labels_==0])


# In[8]:


draw_fruits(fruits[km.labels_==1])


# In[9]:


draw_fruits(fruits[km.labels_==2])


# ## 클러스터 중심

# In[10]:


draw_fruits(km.cluster_centers_.reshape(-1, 100, 100), ratio=3)


# In[11]:


print(km.transform(fruits_2d[100:101]))


# In[12]:


print(km.predict(fruits_2d[100:101]))


# In[13]:


draw_fruits(fruits[100:101])


# In[14]:


print(km.n_iter_)


# ## 최적의 k 찾기

# In[15]:


inertia = []
for k in range(2, 7):
    km = KMeans(n_clusters=k, n_init='auto', random_state=42)
    km.fit(fruits_2d)
    inertia.append(km.inertia_)

plt.plot(range(2, 7), inertia)
plt.xlabel('k')
plt.ylabel('inertia')
plt.show()