Add solution 295

Author: yanglbme

README.md

@@ -70,6 +70,7 @@ Complete solutions to Leetcode problems, updated daily.
 | 032 | [Longest Valid Parentheses](https://github.com/doocs/leetcode/tree/master/solution/032.Longest%20Valid%20Parentheses) | `String`, `Dynamic Programming` |
 | 084 | [Largest Rectangle in Histogram](https://github.com/doocs/leetcode/tree/master/solution/084.Largest%20Rectangle%20in%20Histogram) | `Array`, `Stack` |
 | 145 | [Binary Tree Postorder Traversal](https://github.com/doocs/leetcode/tree/master/solution/145.Binary%20Tree%20Postorder%20Traversal) | `Stack`, `Tree` |
+| 295 | [Find Median from Data Stream](https://github.com/doocs/leetcode/tree/master/solution/295.Find%20Median%20from%20Data%20Stream) | `Heap`, `Design` |
 
 ## Contributions
 I'm looking for long-term contributors/partners to this repo! Send me PRs if you're interested! See the following:

solution/295.Find Median from Data Stream/README.md

@@ -0,0 +1,81 @@
+## 数据流的中位数
+### 题目描述
+
+中位数是有序列表中间的数。如果列表长度是偶数，中位数则是中间两个数的平均值。
+
+例如，
+
+[2,3,4] 的中位数是 3
+
+[2,3] 的中位数是 (2 + 3) / 2 = 2.5
+
+设计一个支持以下两种操作的数据结构：
+
+- void addNum(int num) - 从数据流中添加一个整数到数据结构中。
+- double findMedian() - 返回目前所有元素的中位数。
+
+示例：
+```
+addNum(1)
+addNum(2)
+findMedian() -> 1.5
+addNum(3) 
+findMedian() -> 2
+```
+
+进阶:
+
+- 1. 如果数据流中所有整数都在 0 到 100 范围内，你将如何优化你的算法？
+- 2. 如果数据流中 99% 的整数都在 0 到 100 范围内，你将如何优化你的算法？
+
+### 解法
+维护一个大根堆 bigRoot 和一个小根堆 smallRoot，若有 n 个数，较小的 n/2 个数放在大根堆，而较大的 n/2 个数放在小根堆。
+
+- 若其中一个堆的元素个数比另一个堆的元素个数大 1，弹出堆顶元素到另一个堆。
+- 获取中位数时，若两个堆的元素个数相等，返回两个堆堆顶的平均值。否则返回元素个数较多的堆的堆顶。
+
+```java
+class MedianFinder {
+    
+    private PriorityQueue<Integer> bigRoot;
+    private PriorityQueue<Integer> smallRoot;
+
+    /** initialize your data structure here. */
+    public MedianFinder() {
+        bigRoot = new PriorityQueue<>(Comparator.reverseOrder());
+        smallRoot = new PriorityQueue<>(Integer::compareTo);
+    }
+    
+    public void addNum(int num) {
+        if (bigRoot.isEmpty() || bigRoot.peek() > num) {
+            bigRoot.add(num);
+        } else {
+            smallRoot.add(num);
+        }
+        
+        int size1 = bigRoot.size();
+        int size2 = smallRoot.size();
+        if (size1 - size2 > 1) {
+            smallRoot.add(bigRoot.poll());
+        } else if (size2 - size1 > 1) {
+            bigRoot.add(smallRoot.poll());
+        }
+    }
+    
+    public double findMedian() {
+        int size1 = bigRoot.size();
+        int size2 = smallRoot.size();
+        
+        return size1 == size2 
+            ? (bigRoot.peek() + smallRoot.peek()) * 1.0 / 2
+            : (size1 > size2 ? bigRoot.peek() : smallRoot.peek());
+    }
+}
+
+/**
+ * Your MedianFinder object will be instantiated and called as such:
+ * MedianFinder obj = new MedianFinder();
+ * obj.addNum(num);
+ * double param_2 = obj.findMedian();
+ */
+```

solution/295.Find Median from Data Stream/Solution.java

@@ -0,0 +1,41 @@
+class MedianFinder {
+
+    private PriorityQueue<Integer> bigRoot;
+    private PriorityQueue<Integer> smallRoot;
+
+    /** initialize your data structure here. */
+    public MedianFinder() {
+        bigRoot = new PriorityQueue<>(Comparator.reverseOrder());
+        smallRoot = new PriorityQueue<>(Integer::compareTo);
+    }
+
+    public void addNum(int num) {
+        if (bigRoot.isEmpty() || bigRoot.peek() > num) {
+            bigRoot.add(num);
+        } else {
+            smallRoot.add(num);
+        }
+
+        int size1 = bigRoot.size();
+        int size2 = smallRoot.size();
+        if (size1 - size2 > 1) {
+            smallRoot.add(bigRoot.poll());
+        } else if (size2 - size1 > 1) {
+            bigRoot.add(smallRoot.poll());
+        }
+    }
+
+    public double findMedian() {
+        int size1 = bigRoot.size();
+        int size2 = smallRoot.size();
+
+        return size1 == size2 ? (bigRoot.peek() + smallRoot.peek()) * 1.0 / 2
+                : (size1 > size2 ? bigRoot.peek() : smallRoot.peek());
+    }
+}
+
+/**
+ * Your MedianFinder object will be instantiated and called as such:
+ * MedianFinder obj = new MedianFinder(); obj.addNum(num); double param_2 =
+ * obj.findMedian();
+ */