各种聚类算法的使用对比

ゞ浴缸里的玫瑰 2023-07-02 11:25 90阅读 0赞

#### 1. 导入数据 ####

# beer dataset
    import pandas as pd
    beer = pd.read_csv('data.txt', sep=' ')
    beer

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70]

### 2.构建标签 ###

X = beer[["calories","sodium","alcohol","cost"]]

### 3.使用聚类算法 ###

#### K-means clustering： ####

from sklearn.cluster import KMeans
    
    km = KMeans(n_clusters=3).fit(X)
    km2 = KMeans(n_clusters=2).fit(X)
    
    km.labels_

![在这里插入图片描述][20200129165927360.png]

beer['cluster'] = km.labels_
    beer['cluster2'] = km2.labels_
    beer.sort_values('cluster')

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 1]

from pandas.plotting import scatter_matrix
    %matplotlib inline
    
    cluster_centers = km.cluster_centers_
    
    cluster_centers_2 = km2.cluster_centers_

beer.groupby("cluster").mean()

![在这里插入图片描述][20200130104117572.png]

beer.groupby("cluster2").mean()

![在这里插入图片描述][20200130104133577.png]

centers = beer.groupby("cluster").mean().reset_index()

%matplotlib inline
    import matplotlib.pyplot as plt
    plt.rcParams['font.size'] = 14

import numpy as np
    colors = np.array(['red', 'green', 'blue', 'yellow'])

plt.scatter(beer["calories"], beer["alcohol"],c=colors[beer["cluster"]])
    
    plt.scatter(centers.calories, centers.alcohol, linewidths=3, marker='+', s=300, c='black')
    
    plt.xlabel("Calories")
    plt.ylabel("Alcohol")

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 2]

scatter_matrix(beer[["calories","sodium","alcohol","cost"]],s=100, alpha=1, c=colors[beer["cluster"]], figsize=(10,10))
    plt.suptitle("With 3 centroids initialized")

注释

scatter_matrix(frame, alpha=0.5, c,figsize=None, ax=None, diagonal='hist', marker='.', density_kwds=None,hist_kwds=None, range_padding=0.05, **kwds)

参数含义

1.  frame:pandas dataframe对象；
2.  alpha:图像透明度
3.  figsize:以英寸为单位的图像大小，一般以元组 (width, height) 形式设置
4.  ax:可选一般为none
5.  diagonal:必须且只能在\{‘hist’, ‘kde’\}中选择1个，’hist’表示直方图(Histogram plot),’kde’表示核密度估计(Kernel Density Estimation)；该参数是scatter\_matrix函数的关键参数
6.  marker:Matplotlib可用的标记类型，如’.’，’,’，’o’等
7.  density\_kwds:(other plotting keyword arguments，可选)，与kde相关的字典参数
8.  hist\_kwds:与hist相关的字典参数
9.  range\_padding:(float, 可选)，图像在x轴、y轴原点附近的留白(padding)，该值越大，留白距离越大，图像远离坐标原点
10. kwds:与scatter\_matrix函数本身相关的字典参数  
    ![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 3]

scatter_matrix(beer[["calories","sodium","alcohol","cost"]],s=100, alpha=1, c=colors[beer["cluster2"]], figsize=(10,10))
    plt.suptitle("With 2 centroids initialized")

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 4]

### Scaled data ###

from sklearn.preprocessing import StandardScaler
    scaler = StandardScaler()
    X_scaled = scaler.fit_transform(X)
    X_scaled

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 5]

km = KMeans(n_clusters=3).fit(X_scaled)

beer["scaled_cluster"] = km.labels_
    beer.sort_values("scaled_cluster")

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 6]

beer.groupby("scaled_cluster").mean()

![aidu][]

pd.plotting.scatter_matrix(X, c=colors[beer.scaled_cluster], alpha=1, figsize=(10,10), s=100)

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 7]

## 聚类评估：轮廓系数（Silhouette Coefficient ） ##

s ( i ) = b ( i ) − a ( i ) m a x \{ a ( i ) , b ( i ) \} , s ( i ) = \{ 1 − a ( i ) b ( i ) , a ( i ) < b ( i ) 0 , a ( i ) = b ( i ) b ( i ) a ( i ) − 1 , a ( i ) > b ( i ) \} s(i)=\\frac\{b(i)-a(i)\}\{max\\\{a(i),b(i)\\\}\},s(i)=\\begin\{Bmatrix\} 1-\\frac\{a(i)\}\{b(i)\} & ,a(i)<b(i)\\\\ 0 & ,a(i)=b(i)\\\\ \\frac\{b(i)\}\{a(i)\}-1 & ,a(i)>b(i) \\end\{Bmatrix\} s(i)=max\{ a(i),b(i)\}b(i)−a(i),s(i)=⎩⎪⎨⎪⎧1−b(i)a(i)0a(i)b(i)−1,a(i)<b(i),a(i)=b(i),a(i)>b(i)⎭⎪⎬⎪⎫

*  计算样本i到同簇其他样本的平均距离ai。ai 越小，说明样本i越应该被聚类到该簇。将ai 称为样本i的簇内不相似度。
 *  计算样本i到其他某簇Cj 的所有样本的平均距离bij，称为样本i与簇Cj 的不相似度。定义为样本i的簇间不相似度：bi =min\{bi1, bi2, …, bik\}

*  si接近1，则说明样本i聚类合理
 *  si接近-1，则说明样本i更应该分类到另外的簇
 *  若si 近似为0，则说明样本i在两个簇的边界上。

from sklearn import metrics
    score_scaled = metrics.silhouette_score(X,beer.scaled_cluster)
    score = metrics.silhouette_score(X,beer.cluster)
    print(score_scaled, score)

![在这里插入图片描述][20200206161747661.png]

scores = []
    for k in range(2,20):
        labels = KMeans(n_clusters=k).fit(X).labels_
        score = metrics.silhouette_score(X, labels)
        scores.append(score)
    
    scores

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 8]

plt.plot(list(range(2,20)), scores)
    plt.xlabel("Number of Clusters Initialized")
    plt.ylabel("Sihouette Score")

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 9]

## DBSCAN clustering ##

from sklearn.cluster import DBSCAN
    db = DBSCAN(eps=10, min_samples=2).fit(X)

labels = db.labels_

beer['cluster_db'] = labels
    beer.sort_values('cluster_db')

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 10]

beer.groupby('cluster_db').mean()

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 11]

pd.plotting.scatter_matrix(X, c=colors[beer.cluster_db], figsize=(10,10), s=100)

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 12]

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/2020012916552256.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[20200129165927360.png]: https://img-blog.csdnimg.cn/20200129165927360.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20200129170745906.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[20200130104117572.png]: https://img-blog.csdnimg.cn/20200130104117572.png
[20200130104133577.png]: https://img-blog.csdnimg.cn/20200130104133577.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20200201163741254.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 3]: https://img-blog.csdnimg.cn/20200201165122470.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 4]: https://img-blog.csdnimg.cn/20200205151536259.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 5]: https://img-blog.csdnimg.cn/20200205204623650.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 6]: https://img-blog.csdnimg.cn/20200205204758620.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[aidu]: https://img-blog.csdnimg.cn/20200205204817457.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 7]: https://img-blog.csdnimg.cn/20200205205233169.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[20200206161747661.png]: https://img-blog.csdnimg.cn/20200206161747661.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 8]: https://img-blog.csdnimg.cn/20200206161817292.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 9]: https://img-blog.csdnimg.cn/20200206161855668.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 10]: https://img-blog.csdnimg.cn/2020020616204832.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 11]: https://img-blog.csdnimg.cn/20200206162125911.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg_size_16_color_FFFFFF_t_70 12]: https://img-blog.csdnimg.cn/20200206162216493.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3dlaXhpbl80MjI2MDEwMg==,size_16,color_FFFFFF,t_70