How to set initial size for HashMap

1. Requirements of Alibaba code specifications

2. When scanning using Alibaba plug-in

3. Source code

3.1 When initialization does not set the size

Map<Integer, BigDecimal> staffDidiMap = new HashMap<>();

3.2 When initializing the size

Map<Integer, BigDecimal> staffDidiMap = new HashMap<>(16);

@Test
	public void caffineTest() {
    
    
		System.out.println(1 << 30);
	}
result: 1073741824

---

this.threshold = tableSizeFor(initialCapacity);

@Test
	public void caffineTest() {
    
    
		int cap = 16;
		int n = cap - 1;
		n |= n >>> 1;
		n |= n >>> 2;
		n |= n >>> 4;
		n |= n >>> 8;
		n |= n >>> 16;
		int result = (n < 0) ? 1 : (n >= 1073741824) ? 1073741824 : n + 1;
		System.out.println(result);
	}
result: 16

//TODO source code part needs to be improved

4. Test

@SpringBootTest
public class CaffineTest {
    
    

	@Test
	public void caffineTest() throws InterruptedException {
    
    
		test1();
		test2();
		test3();
		test4();
		test5();
	}

	void test1() throws InterruptedException {
    
    
		int aHundredMillion = 10000000;
		Map<Integer, Integer> map = new HashMap<>();
		long s1 = System.currentTimeMillis();
		for (int i = 0; i < aHundredMillion; i++) {
    
    
			map.put(i, i);
		}
		long s2 = System.currentTimeMillis();
		System.out.println("未初始化容量，耗时 ： " + (s2 - s1));
		map.clear();
		Thread.sleep(5000);
	}

	void test2() throws InterruptedException {
    
    
		int aHundredMillion = 10000000;
		Map<Integer, Integer> map1 = new HashMap<>(aHundredMillion / 2);
		long s5 = System.currentTimeMillis();
		for (int i = 0; i < aHundredMillion; i++) {
    
    
			map1.put(i, i);
		}
		long s6 = System.currentTimeMillis();
		System.out.println("初始化容量5000000，耗时 ： " + (s6 - s5));
		map1.clear();
		Thread.sleep(5000);
	}

	void test3() throws InterruptedException {
    
    
		int aHundredMillion = 10000000;
		Map<Integer, Integer> map2 = new HashMap<>(aHundredMillion);
		long s3 = System.currentTimeMillis();
		for (int i = 0; i < aHundredMillion; i++) {
    
    
			map2.put(i, i);
		}
		long s4 = System.currentTimeMillis();
		System.out.println("初始化容量为10000000，耗时 ： " + (s4 - s3));
		map2.clear();
		Thread.sleep(5000);
	}

	void test4() throws InterruptedException {
    
    
		int aHundredMillion = 10000000;
		// 按照阿里巴巴规范   initialCapacity = (需要存储的元素个数 / 负载因子) + 1。注意负载因子（即 loader factor）默认 为 0.75
		int initial = (int) (aHundredMillion / 0.75F + 1F);
		Map<Integer, Integer> map3 = new HashMap<>(initial);
		long s7 = System.currentTimeMillis();
		for (int i = 0; i < aHundredMillion; i++) {
    
    
			map3.put(i, i);
		}
		long s8 = System.currentTimeMillis();
		System.out.println("初始化容量为" + initial + "耗时" + (s8 - s7));
		map3.clear();
		Thread.sleep(5000);
	}

	void test5() throws InterruptedException {
    
    
		int aHundredMillion = 10000000;
		// 初始化16
		Map<Integer, Integer> map4 = new HashMap<>(16);
		long s9 = System.currentTimeMillis();
		for (int i = 0; i < aHundredMillion; i++) {
    
    
			map4.put(i, i);
		}
		long s10 = System.currentTimeMillis();
		System.out.println("初始化容量为" + 16 + "耗时" + (s10 - s9));
		map4.clear();
		Thread.sleep(5000);
	}
}

Result:

Ten million pieces of data are inserted into the map each time.
Method 1: Not initialized,
time consuming 3995. Method 2: Initialization to 5 million,
time consuming 2345. Method 3: Initialization to 10 million
, time consuming 261. Method 4: Initialization according to Alibaba specifications, time consuming 192.
Method 5: Initialization 16, time consuming 182.

! ! !
After many tests, it can be concluded that the worst performance is uninitialized, the most efficient and stable one is according to Alibaba specifications initialCapacity = (需要存储的元素个数 / 负载因子) + 1。注意负载因子（即 loader factor）默认 为 0.75, then 16 and finally according to the size of the assigned size.

appendix

1. Why does Alibaba recommend that the capacity of HashMap needs to be specified when initializing it?