Create vector.py

Author: Sweetlemon68

cyaron/vector.py

@@ -0,0 +1,71 @@
+from .consts import ALPHABET_SMALL
+from .utils import *
+import random
+
+
+class Vector:
+    @staticmethod
+    def random(num=5, position_range=[10], mode=0, **kwargs):
+        # mode 0=unique 1=repeatable 2=float
+        if(num > 1000000):
+            raise Exception("num no more than 1e6")
+        if(not list_like(position_range)):
+            raise Exception("the 2nd param must be a list, whose length = 1st param")
+        dimension = len(position_range)
+        offset = []
+        vector_space = 1
+        for i in range(0, dimension):
+            if(list_like(position_range[i])):
+                if(position_range[i][1] < position_range[i][0]):
+                    raise Exception("max should larger than min")
+                offset.insert(i, position_range[i][0])
+                position_range[i] = position_range[i][1] - offset[i]
+            else:
+                offset.insert(i, 0)
+            if(position_range[i] <= 0):
+                raise Exception("the difference must more than 0")
+            vector_space *= (position_range[i] + 1)
+        result = []
+        
+        if(mode == 2 or mode == 1):
+            for i in range(0, num):
+                tmp = []
+                for j in range(0, dimension):
+                    one_num = random.randint(0,position_range[j]) if mode == 1 else random.uniform(0,position_range[j])
+                    tmp.insert(j, one_num + offset[j])
+                result.insert(i, tmp)
+                
+        elif((mode == 0 and vector_space > 5 * num)):
+            num_set = set([])
+            rand = 0;
+            for i in range(0, num):
+                while True:
+                    rand = random.randint(0, vector_space - 1);
+                    if(not rand in num_set):
+                        break
+                # Todo: 这边效率如何？我认为是logn级别的检查
+                num_set.add(rand)
+                tmp = Vector.get_vector(dimension, position_range, rand)
+                for j in range(0, dimension):
+                    tmp[j] += offset[j]
+                result.insert(i, tmp)
+                         
+            
+        else:
+            # 生成0~vector_space的所有向量空间
+            rand_arr = [i for i in range(0, vector_space)]
+            random.shuffle(rand_arr)
+            for i in range(0, num):
+                tmp = Vector.get_vector(dimension, position_range, rand_arr[i])
+                for j in range(0, dimension):
+                    tmp[j] += offset[j]
+                result.insert(i, tmp)
+        return result
+
+    @staticmethod      
+    def get_vector(dimension, position_range, hashnum):
+        tmp = []
+        for i in range(0, dimension):
+            tmp.insert(i, hashnum % (position_range[i] + 1))
+            hashnum //= (position_range[i] + 1)
+        return tmp