Trapping Rain Water between Towers

How Much Water Can A Bar Graph Hold 

---

Given an bar graph, encoded as an array of non-negative integers, calculate the units of water that this bar graph can hold.

Let us put in technical terms,
Given n non-negative integers representing an elevation map where the width of each bar is 1, compute how much water it is able to trap after raining. Lets understand the problem statement graphically and it will be more clear,

Case 1:
Input     = [2, 0, 2]
Output  = 2


Case 2:
Input     = [3, 0, 0, 2, 0, 4]
Output  = 10

Case 3:
Input     = [0, 1, 0, 2, 1, 0, 1, 3, 2, 1, 2, 1]
Output  =  6

Read more

How ConcurrentHashMap works and ConcurrentHashMap interview questions.

ConcurrentHashMap Interview Questions In Java.

---

Question 1.  What is the need of ConcurrentHashMap when there is HashMap and Hashtable already present?

Performance and Thread safety are 2 parameter on which ConcurrentHashMap is focused.
Imagine a scenario where we have frequent reads(get) and less writes(put) and need thread safety,

Can we use Hashtable in this scenario?
No. Hashtable is thread safe but give poor performance in case of multiple thread reading from hashtable because all methods of Hashtable including get() method is synchronized and due to which invokation to any method has to wait until any other thread working on hashtable complete its operation(get, put etc).

Can we use HashMap in this scenario?
No. Hashmap will solve performance issue by giving parallel access to multiple threads reading hashmap simultaneously.
But Hashmap is not thread safe, so what will happen if one thread tries to put data and requires Rehashing and at same time other thread tries to read data from hashmap, It will go in infinite loop.

Infinite loop problem discussed in detail: Infinite loop in HashMap
 

ConcurrentHashMap combines good features of hashmap and hashtable and solves performance and thread safety problem nicely.

---

Question 2.  HashMap and Hashtable uses Array and Linkedlist as datastructure to store data, How is it different in ConcurrentHashMap?

If you are not familiar with HashMap and Hashtable, Please go through it first:
How HashMap works

Below diagram shows how hashtable/hashmap look like,

Hashmap internal data structure

ConcurrentHashMap added one Array on top of it and each index of this additional array represents complete HashMap. Additional array is called Segment in ConcurrentHashMap.

Architecture of ConcurrentHashMap looks like below,

 Internal structure of ConcurrentHashMap

Putting key-value pair: 

1. Putting key-value pair in ConcurrentHashMap requires first identifying exact index in 
    Segment array. 
    (Once Segment array index is identified, Now flow will be exactly same as putting the data in 
    hashmap/hashtable.) 
2. After identifying index in Segment array, next task is to identify index of internal bucket/array 
    present in internal hashmap as shown in figure above.
 3. After identying bucket(internal array index), iterate key-value pairs and check each key with key 
    to store, wherever match is found replace stored value with value to store.
    If there is no match, store key-value pair at the last of list.


Getting key-value pair:

1. Getting key-value pair in ConcurrentHashMap requires first identifying exact index in 
    Segment array. 
    (Once Segment array index is identified, Now flow will be exactly same as getting the data from
    hashmap/hashtable.) 

2. After identifying index in Segment array, next task is to identify index of internal bucket/array 
    present in internal hashmap as shown in figure above.
3. After identying bucket(internal array index), iterate key-value pairs and match each key with 
    given key, wherever match is found return value stored against key.
    If there is no match, return null.

---

Question 3.  How ConcurrentHashMap is efficient in terms of Performance and Thread safety?

ConcurrentHashMap provides better Performance by replacing the Hashtable's map wide lock to Segment level lock.
Hashtable is not efficient beacause it uses map wide lock, it means lock is applied on map object itself, 

So if 2 threads tries to call hashtable.get(key), 
Thread T1 calls to get() method will acquire a lock on hashtable object and then execute get() method. (Lock is on complete 'hashtable object')

Now if Thread T2 calls hashtable.get(key) method, then it will also try to acquire lock on hashtable object, but T2 will not able to acquire lock as lock on 'hashtable' is currently held by T1,  So T2 waits until T1 finishes get() operation and release lock on hashtable object.

Hashtable vs ConcurrentHashMap

ConcurrentHashMap works bit different here and instead of locking complete map object it Locks per Segment. 
It means instead of single map wide lock, it has multiple Segment level lock.

So 2 Threads can execute put operation simultaneously by acquiring lock on different Segments.
Thread T1 calls concurrentHashMap.put(key, value), It acquires lock on say Segment 1 and invokes put method.
Thread T2 calls concurrentHashMap.put(key, value), It acquires lock on say Segment 4 and invokes put method.

Both threads doesn't interfere with each other and both can proceed simultaneously as they are working on separate Segment locks.

This is how ConcurrentHashMap improves Performance and provide Thread safety as well. 

---

Question 4.  Can multiple threads read and write from same or different Segments of ConcurrentHashMap simultaneously?

Read Operation: get(key)
Same Segment/Different Segment : Yes. 
Two threads T1 and T2 both can simultaneously read data from same Segment or different Segment of CHM simultaneously without blocking each other.


Write Operation: put(key, value)
Different Segment :Yes
Multiple threads can write data to different Segment of CHM simultaneously without blocking each other.
Same Segment : No  
Multiple threads CANNOT write data to same Segment of CHM simultaneously and need to wait for one thread to come write operation and then only other write operation can be proceed.
  

Read-Write Operation: get and put
Say T1 is writing data in Segment 1 and T2 is reading data from same Segment 1, can read be allowed while writing operation is going on?
YES. 
Both operation that is T1 writing and T2 reading can be done parallely. 

What data will T2 read if T1 is updating same data?
Retrieval operations (including get) generally do not block, so may overlap with update operations (including put and remove). 
Latest updated value present will be returned by get operation that is value updated by most recently completed update operationswill be returned.

Note: Get operations are lock free and can be performed simulateneously irrespective of other thread writing data on same or different Segment of CHM.

---

Question 5.  What is the default size of Segment array? how it is tuned? What is ConcurrenyLevel in case of CHM?

Default size of Segment array is 16. 
 
ConcurrentHashMap differes from Hashtable in terms of Performance by introducing Segment array.
Each index of Segment array is guarded by a lock for put operation.

Concurrency level in ConcurrentHashmap 

Threads working on separate Segments index doesn't affect each other. 
By default Segments array size is 16, So maximum 16 threads can simultaneously put data in map considering each thread is working on separate Segment array index.

How Segment array size is tuned?

Segment size decides the number of Threads that can paralley write to a map.
Segment array size is configured using ConcurrencyLevel parameter as shown below,

ConcurrentHashMap m = new ConcurrentHashMap(initialCapacity, loadFactor, concurrencyLevel)

Concurrenthashmap segment vs Hashmap bucket

It takes 3 parameters,   

Example:

ConcurrentHashMap m = new ConcurrentHashMap(200 , 0.75f, 10);

Initial capacity is 200, it means CHM make sure it has space for adding 200 key-value pairs after creation.

Load factor is 0.75, it means when average number of elements per map exceeds 150 (intital capacity * load factor = 200 * 0.75 = 150) at that time map size will be increased and existing items in map are rehashed to put in new larger size map.
For more details on Load Factor: Load factor in Map

Concurrency level is 10, it means at any given point of time Segment array size will be 10 or greater than 10, so that 10 threads can able to parallely write to a map.

---

Question 6.  What will happen if the size of Segment array is too small or too large?

Choosing correct ConcurrencyLevel is very important because ConcurrencyLevel decides what will be the size of Segment array.

Segment array size will decide how many parallel Threads will be able to execute put operation on map parallely.

Concurrenthashmap should not be too low as well as not too high

So Segment array size should not be too big or should not be too small because,
Using a significantly higher value than we will waste space and time, and a significantly lower value can lead to thread competition.

---

Question 6.  If we choose ConcurrenyLevel as 10 then what will be size of Segment array? Is Segment array size exactly same as concurrenyLevel? If No, then how is the Segment array size calculated?

Segment array size is calculated based on concurrenyLevel specified but it doesn't mean it will be exactly same as concurrenyLevel.

If concurrenyLevel is 10 then Segment array size will be 16.

Segment array size = 2 to the power x, where result should be >= concurrenyLevel(in our case it is 10)
Segment array size = 2 to the power x >= 10

Segment array size = 2 ^ 1 = 2   >= 10 (False)
Segment array size = 2 ^ 2 = 4   >= 10 (False)
Segment array size = 2 ^ 3 = 8   >= 10 (False)
Segment array size = 2 ^ 4 = 16 >= 10 (True) 

Segment array size is 16.

Example: 2
concurrenyLevel = 8 then Segment array size = ?
Find 2 ^ x >= 8 

2 ^ 1 >= 2 
2 ^ 2 >= 4
2 ^ 3 >= 8
Segment array size will be 8.

---

Question 7.  What is HashEntry array and how is the size of HashEntry decided?

Default initial capacity of CHM is 16. It means CHM make sure there is sufficient space to accomodate 16 key-value pairs after CHM is created.


What is HashEntry[]?

We saw that each index of Segment array itself is complete HashMap, 

Hashentry array in Concurrenthashmap

 In HashMap the bucket/array is of class Entry[] and in CHM the array is of class HashEntry[].
 

static final class Segment<K,V> extends ReentrantLock implements Serializable {
 transient volatile HashEntry<K,V>[] table;
}

static final class HashEntry<K,V> {
 final int hash;
 final K key;
 volatile V value;
 volatile HashEntry<K,V> next;
}

Lets see how internal architecture of CHM looks like, 

How concurrenthashmap works internally in java

How HashEntry[] array size will is calculated?

ConcurrentHashMap m = new ConcurrentHashMap(initialCapacity, loadFactor, concurrencyLevel)

HashEntry[] array size  =   2 ^ x   >=  (initialCapacity / concurrenyLevel)

Eg: ConcurrentHashMap(32,   0.75f,   4);
HashEntry[] array size  =  2 ^ 1 = 2   >=  8(32/4) (False)
HashEntry[] array size  =  2 ^ 2 = 4   >=  8 (False)
HashEntry[] array size  =  2 ^ 3 = 8   >=  8 (True)

HashEntry[] array size = 8. 
It means there will always be capacity of 8 key-value pairs that can be put in CHM after its creation.

---

Question 8.  How does ConcurrentHashMap handle rehashing while another thread is still writing on another segment/partition?

For understanding rehashing, please understand Load Factor first.
Load Factor is a measure, which decides when exactly to increase the HashMap/CHM capacity(buckets) to maintain get and put operation complexity of O(1).

Default load factor of Hashmap/CHM is 0.75f (i.e 75% of current map size).

For more details on Load factor, Please refer: Load Factor in HashMap

How rehashing works in Concurrenthashmap

In CHM, Every segment is separately rehashed so there is no collision between Thread 1 writing to Segment index 2 and Thread 2 writing to Segment index 5.

Example: If say Thread 1 which is putting data in Segment[] array index 2 finds that HashEntry[] array needs to be rehashed due to exceed Load factor capacity then it will rehash HashEntry[] array present at Segment[] array index 2 only.
HashEntry[] array at other Segment indexes will still be intact, unaffected and continue to serve put and get request parallel.

---

Exception Handling Interview Question-Answer

Method Overloading - Method Hiding Interview Question-Answer

Advanced Multithreading Interview Questions-Answers In Java

Type Casting Interview Questions-Answers In Java

How Thread.join() in Java works internally

How is ambiguous overloaded method call resolved in java

 

Enjoy !!!! 

If you find any issue in post or face any error while implementing, Please comment.

Read more

Java Interface interview questions and answers

Interface Interview Questions In Java.

---

Interview questions on Java Interface

Question 1. Super class of all classes is java.lang.Object class, does this applies to Interface as well? What is super class of all Interface in Java?

java.lang.Object class is the super class of all Java classes. All non-primitive types (including arrays) inherit either directly or indirectly from Object class. 

For Interface that is not the case, Super class of Interface is null.


From below program that will be clear.

interface ITest{}
class CTest{}

class Test{ 
 public static void main(String[] args) {
  System.out.println(CTest.class.getSuperclass()); // class java.lang.Object
  System.out.println(int[].class.getSuperclass()); // class java.lang.Object
  System.out.println(ITest.class.getSuperclass()); // null
  System.out.println(int.class.getSuperclass());   // null
 }
}

Output:
 class java.lang.Object
 class java.lang.Object
 null
 null

Read more

The Knapsack Problem

0/1 Knapsack Problem solved using Iterative and Dynamic Programming

---

Given weights and values(profits) of N items, put these items in a knapsack of max capacity W to get the maximum total value(profit) in the knapsack.

Read more

Advanced Multithreading Interview Questions In Java

Multithreading Tutorial.
Interview Questions on Threads in Java

---

Question 1. What is the use of Threads in Java? why Thread is required? What are Threads in Java?

Lets try to understand this with simple scenario and it will be more clear: 

Scenario:
Suppose you want to count the population of a India, how will you approach? 
Note: There are 29 states in India. 

Approach 1:
First approach is, you start with first state and count population of that state then you will start second state and so on for all 29 states.
Once you have population of all the states, just sum the population count of all States. 

Imagine the time it will take for you to do this as you are alone and you have to count population state by state. 


Approach 2:
Second approach is, you called 29 people to help you out and you distributed the task of population count to 29 person, each person taking care of individual state. 
So, person 1 will take care of population count for State 1. 
Person 2 will take care of population count for State 2 and so on. 

Once you have population count of all the states, just sum the population count received from all 29 person. 

Imagine the time it will take for you to do this as compared to Approach 1, surely it will be much less. 
   
So that is what Thread does. In above scenario, you can consider 29 persons as 29 Threads who are doing their respective task of population count. 
It is possible that Person 1 may finish population count for State 1 assigned to it much early than Person 2 doing population count for State 2 because State 1 was small. 
Person 2 will continue doing his task even after Person 1 finished early. 

In the similar way, Say If you have 2 Threads say Thread 1 and Thread 2. Thread 1 may complete its job early and Thread 2 will continue doing its job even after Thread 1 is done and they both execute separately. 

Now to relate it with Threads:
When you have task like above that needs to be run in parallel for faster processing at that time Threading will come in picture.

You can say, Java Threads helps creating multiple independent path of execution within a program which can run parallely.

Application Example:
In Java, when a program requires more than one task to execute in parallel, say for example,

1. Reading a data from a local file.
2. Reading a data from remote connection.
   

When both of above task need to be executed in parallel at that time Threading will come in picture.
So Java Threads helps creating multiple independent path of execution within a program which can run in parallel.  

Read more

Method Overriding Rules In Java

Method overriding rules Java

---

Parameters that need to consider in method overriding is,

What is Method Overriding?

In case of Inheritance, where Parent-child relationship exist, 
Parent class define a method say connect() which does connection to some remote service. 
Child class has better way to do connection than Parent class and don't want to use Parent class connect() method, So it overrides Parent's class connect() method by providing own implementation of connect() method, 

Now whenever connect() method is called using Child class object, then connect() method of Child class will be called and not the connect() method of Parent/Super class, this concept is called Method overriding.

Example of Method Overriding: 

class Parent{
 public void connect(){  
  System.out.println("Doing connection in Parent way");  
 }
}

class Child extends Parent {
 @Override
 public void connect() { 
  System.out.println("Doing connection in Child way");
 }
}

Rules of Method Overriding

---

Parameters that need to be consider in case of Method Overriding,

1. Access Specifier of method
2. Return Type of method
3. Name of method
4. Arguments/Parameters of method
5. Exception that method throws. 

Access Specifier of method
Overriding method in Child class can either have same access specifier as that of Parent class method or can increase visibility but cannot decrease it.

If Parent class method is declared as,

protected void connect(){}

then valid access specifier for Child class overriding method connect() is, 

public void connect(){} OR
protected void connect(){}

Return Type of method:
Overriding method in Child class can either have same return type or should be Subclass of return type declared in method of Super class.

If Parent class method is declared as,

protected Number connect(){}

then valid Return type for overriding connect() method in Child class is either Number class or all subclass of Number class, 

public Integer connect(){}
protected Long connect(){}
protected Number connect(){}

Name of method:
Name of the overriding method in Child class must be exactly same as that of method declared in Parent Class


Arguments/Parameters of method: 
Total number and Type of arguments present in overriding Child class method must be exactly same as that of Parent class method.

Note:
Arguments present in Child class method should be exactly of same type (Subclass will not work) as that of Parent class method.
Why Subclass argument won't work, Let's understand it with below example,

class A1{}
class A2 extends A1{}
class A3 extends A1{}

class Parent{
 protected Number connect(A1 a1){  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }
}

class Child extends Parent {
 @Override
 public Long connect(A2 a2) { 
  System.out.println("Doing connection in Child way");
  return null;
 }
}

In above example,
Super class connect method take generic agument A1.
Child class overriding connect method take specific agument A2

If Covariant parameters are allowed in Method overriding than what will be output of below lines,

Parent p = new Child();
p.connect(new A3());

Above line will call connect() method of Child class method due to Polymorphic behaviour. Also, According to Parent class method definition it should work but according to Child class overriding definition it will not work.

This problem occurred because Child class overriding method accepted specific(Subclass) argument compared to Parent class method which is accepting generic argument.

To avoid this situation, parameter in overriding method must be exactly same. 

Exception that method throws: 

Unchecked Exception:  
Overriding method in Child class can throw any number of Unchecked Exception irrespective of Parent class overriden method has declared any Checked/Unchecked Exception or not.  
Below example shows valid overriden method connect() and connect1().

class Parent{
 protected Object connect(String s1) {  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }

 protected Object connect1(String s1) throws NullPointerException{  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }
}

class Child extends Parent {
 @Override
 public Integer connect(String s2) throws RuntimeException, NullPointerException{ 
  System.out.println("Doing connection in Child way");
  return null;
 }

 @Override
 protected Object connect1(String s1) throws RuntimeException{  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }

 public static void main(String[] args) {
  Parent p = new Child();
  p.connect("hello");
 }
}

Checked Exception:  
If say Overriden method of Parent class throws IOException, then overriding method in Child class can either throw 

1. No Exception,
2. Same IOException,
3. Any number of Subclass of IOException like FileNotFoundException, EOFException etc.

Not Allowed:
If say Overriden method of Parent class throws IOException, then overriding method in Child class cannot throw

1. It cannot throw exception from totally new inheritance hierarchy like SQLException.
2. It cannot throw broader exception like Throwable or Exception in our case. 

Conclusion: 
Overriding method of Child class can throw any number of Unchecked Exception irrespective of Overriden method of Super class throwing or not throwing any exception.

If Super class overriden method throws Cheked Exception then Overriding method of Child class can either choose not to throw any exception, or throws same exception or throws any number of subclass of Exception thrown by overriden method

Lets see few example and understand:

class Parent{
 protected Object connect(String s1) {  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }
 
 protected Object connect1(String s1) throws Exception{  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }

 protected Object connect2(String s1) throws IOException, SQLException{  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }
 
 protected Object connect3(String s1) throws IOException{  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }

 protected Object connect4(String s1){  
  System.out.println("Doing connection in Parent way"); 
  return null;
 }

}

class Child extends Parent {
 @Override
 public Integer connect(String s2){ //It will work
  System.out.println("Doing connection in Child way");
  return null;
 }

 protected Object connect1(String s1) throws Exception, FileNotFoundException, MalformedURLException{ //It will work  
  System.out.println("Doing connection in Child way"); 
  return null;
 }
 
 protected Object connect2(String s1) throws FileNotFoundException{ //It will work
  System.out.println("Doing connection in Child way"); 
  return null;
 }
 
 protected Object connect3(String s1) throws Exception{ //It will NOT work  
  System.out.println("Doing connection in Child way"); 
  return null;
 }
 
 protected Object connect4(String s1) throws Exception{ //It will NOT work  
  System.out.println("Doing connection in Child way"); 
  return null;
 }
 
 public static void main(String[] args) {
  Parent p = new Child();
  p.connect("hello");
 }
}

In case if you like to try program using different Exception hierarchy, here is Exception hierarchy tree.

You may also like to see

---

Advanced Java Multithreading Interview Question-Answer.

How is ambiguous overloaded method call resolved in java?

Exception Handling Interview Question-Answer

Method Overloading - Method Hiding Interview Question-Answer

Type Casting Interview Questions In Java

Interface interview questions and answers in Java

Enjoy !!!! 

If you find any issue in post or face any error while implementing, Please comment.

Read more

Find the number of Islands using DFS.

Count total number of Islands OR
The "Island Count" Problem

---

Given a Ocean in a form of 2D matrix as shown below in which there are few Island present 
(or may not be present).

In a matrix given above, which has only two values ‘1’ and ‘0’. 
1 represents land and 
0 represents water, 
Find the total number of Islands.

Lets understand what is the input and the expected output with few samples.

Sample 1:
Input: 
        int[][] island = new int[][] {
                { 1, 1, 0, 0, 0 },
                { 0, 1, 0, 0, 1 },
                { 1, 0, 0, 1, 1 },
                { 0, 0, 0, 0, 0 },
                { 1, 0, 1, 0, 1 }             
        }; Output: 5

Sample 2:
Input:
        int[][] island = new int[][] {
                {0, 0, 0, 0, 0 },
                {0, 0, 0, 0, 0 },
                {0, 0, 0, 0, 0 },
                {0, 0, 0, 0, 0 },
                {0, 0, 0, 0, 0 }               
        };
Output:0

Sample 3:
Input:
        int[][] island = new int[][] {
                { 1, 0, 0, 0, 0 },
                { 0, 1, 0, 0, 0 },
                { 0, 0, 1, 0, 0 },
                { 0, 0, 0, 0, 0 },
                { 1, 1, 1, 0, 0 }                  
        };
Output:2

Read more

Find the number of Islands

Count number of islands where every island is row-wise and column-wise separated

---

Count total number of islands present in given matrix where every island is separated row-wise and column-wise.

Given a rectangular matrix which has only two values ‘1’ and ‘0’.
1 represents land.
0 represents water.
Islands are all rectangular in shape that is values ‘1’ will always appear in form of rectangular islands and these islands are always row-wise and column-wise separated by at least one line of ‘0’s.

If there is any diagonal island then it will be considered adjacent islands and will be separate.
Count total number of islands in the given matrix.

Lets understand what is the input and the expected output.

Sample 1:
Input: 
        int[][] island = new int[][] {
            { 0,  0,  0 },
            { 1,  1,  0 },
            { 1,  1,  0 },
            { 0,  0,  1 },
            { 0,  0,  1 },
            { 1,  1,  0 }               
        }; Output: 3

Sample 2:
Input:
        int[][] island = new int[][] {
            { 1,  0,  0 },
            { 0,  1,  0 },
            { 0,  0,  1 },
            { 0,  1,  0 },
            { 1,  0,  0 },
            { 0,  1,  0 }               
        };
Output:6

Sample 3:
Input:
        int[][] island = new int[][] {
            { 1,  1,  0 },
            { 1,  1,  0 },
            { 1,  1,  0 },
            { 0,  1,  0 },
            { 0,  0,  1 },
            { 1,  1,  1 }               
        };
Output:3

Read more

Stock Buy Sell to Maximize Profit

Maximum difference between two elements such that larger element appears after the smaller number. OR
Stock Buy Sell to Maximize Profit.

---

We are given an array of N integers representing Stock Prices on a single day.
Find maximum profit that can be earned by performing 1 transaction.

We want to find a pair (buyDay, sellDay), with buyDay ≤ sellDay,
such that if we bought the stock on buyDay and sold it on sellDay, we would maximize our profit.

Lets understand what is the input and the expected output.

Sample 1:
Input: [100, 80, 120, 130, 70, 60, 100, 125]
Output: If we buy a stock at 60 and sell at 125 then profit is maximum (65).
 
Sample 2:
Input: [7, 9, 5, 6, 3, 2]
Output: If we buy a stock at 7 and sell at 9 then profit is maximum (2).

Sample 3:
Input: [6, 5, 4, 3, 2, 1] Output: Prices are in decreasing order so there will be no profit as stock prices goes on decreasing. So answer is 0.

Read more

The Skyline Problem

Skyline Problem In Java.

---

A city's skyline is the outer contour of the silhouette formed by all the buildings in that city when viewed from a distance. 
Now suppose you are given the locations and height of all the buildings as shown on a cityscape below. Write a program to output the skyline formed by these buildings collectively.

Lets simplify the problem statement and understand it correctly,
If there are many buildings in a area as shown in below picture, If same buildings is viewed from distance then what we can see is not all the buildings but the skyline that is borders of all buildings.

You can see skyline of buildings if viewed from a side and remove all sections that are not visible/overlapped.
All buildings have common base and every building is represented by 3 points(left, right, height)

‘left': is x coordinate of building left wall.
‘right': is x coordinate of building right wall
‘height': is height of building.

A skyline is a collection of rectangular strips. A rectangular strip is represented as a pair (left, height) where left is x coordinate of building left wall and height is height of building.

Lets understand what is the input and the expected output. INPUT: You are given a building coordinates as shown below,
int[][] skyscraper = { {2,9,10},  {3,6,15},  {5,12,12},  {13,16,10}, {15,17,5} };

OUTPUT: Skyline Coordinates = { {2,10},  {3,15}, {6,12}, {12,0}, {13,10}, {16,5}, {17,0} }

Read more