Difference between dict implementation principle of Python and HashMap of Java

shzwork 2020-11-13 15:34:29
difference dict implementation principle python

Python It's used in a lot of places inside dict This structure , In object properties __dict__ It's just a dictionary , So it requires high efficiency .

dict Using a hash table , At least in O(1) Time to complete the search . alike java Of HashMap Also uses the hash table implementation , The difference is dict When a hash conflict occurs, it uses Open addressing , and HashMap Adopted Link method .


Open addressing


  1. Records are easier to serialize (serialize) operation
  2. If the total number of records can be predicted , You can create a perfect hash function , The efficiency of processing data is very high


  1. The number of stored records cannot exceed the length of the bucket array , If you exceed it, you need to expand capacity , And expansion can cause the time cost of an operation to skyrocket , This can be a serious drawback in real-time or interactive applications
  2. Using probe sequences , It is possible that the calculated time cost is too high , This results in a decrease in hash table processing performance
  3. Because the records are stored in the bucket array , The bucket array must have empty slots , So when you record your size (size) When it's large and the total number of records is large , The space occupied by empty slots can lead to a significant memory waste
  4. When deleting records , More trouble . Let's say you need to delete a record a, Record b Is in a And then insert the bucket array , But and records a There are conflicts , Is the address found by the probe sequence jumping again , So if you just delete it a,a Becomes an empty slot , An empty slot is the termination condition for a failed query record , This results in a record b stay a Is not visible until the data is reinserted , So you can't just delete it a, Instead, set the delete tag . This requires additional space and operations .
I want to realize one by myself dict Inherit collection Of UserDict, It has encapsulated common methods .
The following is my understanding of how to use python Implemented dictionary , Simplified a lot of functions , For example, the object buffer pool 、String Hash optimization and so on , If there is a wrong or better implementation, please point out . because python There is no pure array structure , So arrays are borrowed list Realized .
from collections import namedtuple
class SimpleArray(object):
# Simple array class implementation 
def __init__(self, mix):
self.container = [None for i in range(mix)]
def __len__(self):
return len(self.container)
def __setitem__(self, key, value):
return self.container.__setitem__(key,value)
def __getitem__(self, item):
return self.container.__getitem__(item)
def __delitem__(self, key):
return self.container.__setitem__(key, None)
def __str__(self):
return str(self.container)
class SimpleDict(object):
# Simple dictionary class implementation 
Init_length = 8 # Initialized size 
Load_factor = 2/3 # Dilatation factor 
def __init__(self):
self._array_len = SimpleDict.Init_length
self._array = SimpleArray(self._array_len)
self._used = 0
self.dictObj = namedtuple("dictObj","key value") # You can use arrays here, or you can use arrays ,namedtuple To make the code more readable 
def __getitem__(self, item):
key = self._hash(item)
dictObj = self._array[key]
if dictObj is not None and dictObj.key == item:
return dictObj.value
for new_key in self._second_hash(key):
if self._array[new_key] is not None and item == self._array[new_key].key:
return self._array[new_key].value
def __setitem__(self, key, value):
# Calculate whether expansion is needed 
if (self._used / self._array_len) > SimpleDict.Load_factor:
# According to the hash Value to calculate the location index 
hash_key = self._hash(key)
new_key = self._second_hash(hash_key)
while True:
if self._array[hash_key] is None or key == self._array[hash_key].key:
# Hash collision occurs. According to the quadratic probe function, the position of the next index is obtained 
hash_key = next(new_key)
if abs(hash_key) >= self._array_len:
hash_key = self._hash(key)
# Find the empty space and put the key value object in 
self._array[hash_key] = self.dictObj(key, value)
self._used += 1
def __delitem__(self, key):
hash_key = self._hash(key)
if key != self._array[hash_key].key:
for new_key in self._second_hash(hash_key):
if key == self._array[new_key].key:
hash_key = new_key
self._array[hash_key] = None
self._used -= 1
def _hash(self, key):
# Calculate the hash value 
return hash(key) & (self._array_len-1)
def _second_hash(self, hash_key):
# Simple quadratic exploration function implementation 
count = 1
for i in range(self._array_len):
new_key = hash_key + count**2
if abs(new_key) < self._array_len:
yield new_key
new_key = hash_key - count**2
if abs(new_key) < self._array_len:
yield new_key
count += 1
def _new_array(self):
# Capacity expansion 
old_array = self._array
self._array_len = self._array_len * 2 # Capacity expansion 2 Multiple size 
self._array = SimpleArray(self._array_len)
for i in range(len(old_array)):
dictObj = old_array[i]
if dictObj is not None:
self[dictObj.key] = dictObj.value
def __str__(self):
result = ", ".join("%s:%s"%(obj.key, obj.value)
for obj in self._array
if obj is not None)
return "{" + result + "}"
if __name__ == '__main__':
d = SimpleDict()
for i in range(20):
d[str(i)] = i
del d["11"]


Link method


  1. For cases where the total number of records changes frequently , It's handled better ( In other words, the overhead of dynamic adjustment is avoided )
  2. Because records are stored in nodes , And nodes are dynamically allocated , It won't waste memory , So it's especially suitable for the size of the record itself (size) A lot of things , Because the cost of the pointer is negligible
  3. When deleting records , It's more convenient , Directly through the pointer operation can be


  1. Stored records are randomly distributed in memory , This way when querying records , Data types that are more compact ( For example, array ), Skip access to the hash table incurs additional time overhead
  2. If all key-value Yes, it can be predicted in advance , And then there is no change ( Insert and delete are not allowed ), You can artificially create a perfect hash function that doesn't conflict (perfect hash function), At this point the performance of closed hashes will be much higher than that of open hashes
  3. Because of the pointer , Records are not easy to serialize (serialize) operation


There are two important parameters that affect its performance : Initial capacity and loading factor

dict: The default initial capacity is 8, The loading factor is 2/3

HashMap: The default initial capacity is 16, The loading factor is 0.75

The same is that the length of expansion must be 2 Of N Power

Edited on 2019-09-04

  1. 利用Python爬虫获取招聘网站职位信息
  2. Using Python crawler to obtain job information of recruitment website
  3. Several highly rated Python libraries arrow, jsonpath, psutil and tenacity are recommended
  4. Python装饰器
  5. Python实现LDAP认证
  6. Python decorator
  7. Implementing LDAP authentication with Python
  8. Vscode configures Python development environment!
  9. In Python, how dare you say you can't log module? ️
  10. 我收藏的有关Python的电子书和资料
  11. python 中 lambda的一些tips
  12. python中字典的一些tips
  13. python 用生成器生成斐波那契数列
  14. python脚本转pyc踩了个坑。。。
  15. My collection of e-books and materials about Python
  16. Some tips of lambda in Python
  17. Some tips of dictionary in Python
  18. Using Python generator to generate Fibonacci sequence
  19. The conversion of Python script to PyC stepped on a pit...
  20. Python游戏开发,pygame模块,Python实现扫雷小游戏
  21. Python game development, pyGame module, python implementation of minesweeping games
  22. Python实用工具,email模块,Python实现邮件远程控制自己电脑
  23. Python utility, email module, python realizes mail remote control of its own computer
  24. 毫无头绪的自学Python,你可能连门槛都摸不到!【最佳学习路线】
  25. Python读取二进制文件代码方法解析
  26. Python字典的实现原理
  27. Without a clue, you may not even touch the threshold【 Best learning route]
  28. Parsing method of Python reading binary file code
  29. Implementation principle of Python dictionary
  30. You must know the function of pandas to parse JSON data - JSON_ normalize()
  31. Python实用案例,私人定制,Python自动化生成爱豆专属2021日历
  32. Python practical case, private customization, python automatic generation of Adu exclusive 2021 calendar
  33. 《Python实例》震惊了,用Python这么简单实现了聊天系统的脏话,广告检测
  34. "Python instance" was shocked and realized the dirty words and advertisement detection of the chat system in Python
  35. Convolutional neural network processing sequence for Python deep learning
  36. Python data structure and algorithm (1) -- enum type enum
  37. 超全大厂算法岗百问百答(推荐系统/机器学习/深度学习/C++/Spark/python)
  38. 【Python进阶】你真的明白NumPy中的ndarray吗?
  39. All questions and answers for algorithm posts of super large factories (recommended system / machine learning / deep learning / C + + / spark / Python)
  40. [advanced Python] do you really understand ndarray in numpy?
  41. 【Python进阶】Python进阶专栏栏主自述:不忘初心,砥砺前行
  42. [advanced Python] Python advanced column main readme: never forget the original intention and forge ahead
  43. python垃圾回收和缓存管理
  44. java调用Python程序
  45. java调用Python程序
  46. Python常用函数有哪些?Python基础入门课程
  47. Python garbage collection and cache management
  48. Java calling Python program
  49. Java calling Python program
  50. What functions are commonly used in Python? Introduction to Python Basics
  51. Python basic knowledge
  52. Anaconda5.2 安装 Python 库(MySQLdb)的方法
  53. Python实现对脑电数据情绪分析
  54. Anaconda 5.2 method of installing Python Library (mysqldb)
  55. Python implements emotion analysis of EEG data
  56. Master some advanced usage of Python in 30 seconds, which makes others envy it
  57. python爬取百度图片并对图片做一系列处理
  58. Python crawls Baidu pictures and does a series of processing on them
  59. python链接mysql数据库
  60. Python link MySQL database