remove numpy

jiagu/cluster/base.py

@@ -1,12 +1,18 @@
 # -*-coding:utf-8-*-
 from collections import Counter
-import numpy as np
+import math
 
 
 def elu_distance(a, b):
-    """计算两点之间的欧氏距离并返回"""
-    dist = np.sqrt(np.sum(np.square(np.array(a) - np.array(b))))
-    return dist
+    """计算两点之间的欧氏距离并返回
+
+    :param a: list of float
+    :param b: list of float
+    :return: float
+    """
+
+    x = sum([pow((a_-b_), 2) for a_, b_ in zip(a, b)])
+    return math.sqrt(x)
 
 
 def count_features(corpus, tokenizer=list):
@@ -15,7 +21,7 @@ def count_features(corpus, tokenizer=list):
     :param corpus: list of str
     :param tokenizer: function for tokenize, default is `jiagu.cut`
     :return:
-        features: np.array
+        features: list of list of float
         names: list of str
 
     example:
@@ -32,7 +38,7 @@ def count_features(corpus, tokenizer=list):
         feature = [counter.get(x, 0) for x in vocab]
         features.append(feature)
 
-    return np.array(features), vocab
+    return features, vocab
 
 
 def tfidf_features(corpus, tokenizer=list):
@@ -41,7 +47,7 @@ def tfidf_features(corpus, tokenizer=list):
     :param corpus: list of str
     :param tokenizer: function for tokenize, default is `jiagu.cut`
     :return:
-        features: np.array
+        features: list of list of float
         names: list of str
 
     example:
@@ -58,9 +64,9 @@ def tfidf_features(corpus, tokenizer=list):
     for word in vocab:
         num = sum([1 if (word in s) else 0 for s in corpus])
         if num == total_doc:
-            idf = np.log(total_doc / num)
+            idf = math.log(total_doc / num)
         else:
-            idf = np.log(total_doc / (num + 1))
+            idf = math.log(total_doc / (num + 1))
         idf_dict[word] = idf
 
     features = []
@@ -69,6 +75,6 @@ def tfidf_features(corpus, tokenizer=list):
         feature = [counter.get(x, 0) / len(sent) * idf_dict.get(x, 0) for x in vocab]
         features.append(feature)
 
-    return np.array(features), vocab
+    return features, vocab
 
 

jiagu/cluster/dbscan.py

@@ -9,7 +9,6 @@
 """
 
 import random
-import numpy as np
 from collections import OrderedDict
 
 from .base import elu_distance
@@ -31,14 +30,13 @@ class DBSCAN(object):
     def train(self, X):
         """输入数据，完成 KMeans 聚类
 
-        :param X: list of tuple / np.array
+        :param X: list of tuple
             输入数据特征，[n_samples, n_features]，如：[[0.36, 0.37], [0.483, 0.312]]
         :return: OrderedDict
         """
-        if isinstance(X, np.ndarray):
-            X = [tuple(x) for x in X.tolist()]
 
         # 确定数据集中的全部核心对象集合
+        X = [tuple(x) for x in X]
         cores = self._find_cores(X)
         not_visit = set(X)
 

jiagu/cluster/kmeans.py

@@ -8,7 +8,6 @@
  * Description : KMeans 算法实现
 """
 
-import numpy as np
 import random
 from collections import OrderedDict
 
@@ -57,11 +56,18 @@ class KMeans(object):
 
         self.clusters = clusters
 
+    def _mean(self, features):
+        res = []
+        for i in range(len(features[0])):
+            col = [x[i] for x in features]
+            res.append(sum(col) / len(col))
+        return res
+
     def _update_centroids(self):
         """根据簇类结果重新计算每个簇的中心，更新 centroids"""
         centroids = []
         for key in self.clusters.keys():
-            centroid = np.mean(self.clusters[key], axis=0)
+            centroid = self._mean(self.clusters[key])
             centroids.append(centroid)
         self.centroids = centroids
 
@@ -88,13 +94,10 @@ class KMeans(object):
     def train(self, X):
         """输入数据，完成 KMeans 聚类
 
-        :param X: list of list / np.array
+        :param X: list of list
             输入数据特征，[n_samples, n_features]，如：[[0.36, 0.37], [0.483, 0.312]]
         :return: OrderedDict
         """
-        if isinstance(X, np.ndarray):
-            X = X.tolist()
-
         # 随机选择 k 个 example 作为初始类簇均值向量
         self.centroids = random.sample(X, self.k)
 

jiagu/cluster/text.py

@@ -5,7 +5,7 @@ from .kmeans import KMeans
 
 
 def text_cluster(docs, features_method='tfidf', method="dbscan",
-				k=3, max_iter=100, eps=0.5, min_pts=2, tokenizer=list):
+                 k=3, max_iter=100, eps=0.5, min_pts=2, tokenizer=list):
     """文本聚类，目前支持 K-Means 和 DBSCAN 两种方法
 
     :param features_method: str
@@ -33,7 +33,7 @@ def text_cluster(docs, features_method='tfidf', method="dbscan",
         raise ValueError('features_method error')
 
     # feature to doc
-    f2d = {k: v.tolist() for k, v in zip(docs, features)}
+    f2d = {k: v for k, v in zip(docs, features)}
 
     if method == 'k-means':
         km = KMeans(k=k, max_iter=max_iter)
@@ -57,7 +57,3 @@ def text_cluster(docs, features_method='tfidf', method="dbscan",
         clusters_out[label] = list(set(c_docs))
 
     return clusters_out
-
-
-
-