Data Structures and Algorithms using Java

Professor Debasis Samanta

Department of Computer Science and Engineering
Indian Institute of Technology Kharagpur
Lecture 9
Map Framework

We are discussing about the collection and then Java supports to facilitate different operations on
the collection. In the last two video lectures we have covered about the different views that a
collection can be seen from the Java programming point of view or Java facilities point of view.
There are many activities, many methods we have learned and altogether we call them as a
collection.

(Refer Slide Time: 1:09)

Today we will discuss another important framework. This is exactly not as a part of Java
collection framework, it is a another framework, more precisely it is called map framework. This
framework like Java collection framework is defined in java.util package. Now it has, it is a,
actually it is a different stock of collection. Different stock of collection means it represents the
objects or rather we can data structures from the different view point. This view point is actually
called map.
(Refer Slide Time: 1:53)

Now in this lecture we will try to understand what are the composition of this map framework
and what are the different classes and interfaces are there, so that map framework can be utilized
by a programmer to develop his or her own application programs.

(Refer Slide Time: 2:23)

Now before going to take the actual discussion, I would like to tell about in what the map
framework is different than the collection framework that we have already studied. Now in the
collection framework the data structure can be viewed either from indexed representation or
sequential representation or linked representation whatever it is there. And there are five
different structures that we have already studied namely arrays, linked list, stack and queue and
set.

Now there are two more structural concepts in the theory of computer science that is there are
two more data structures. They are called tree and table. Somehow we have some idea about tree
based representation of a set. Actually this, in this sense tree structure is used to maintain a
collection in the form of a set and in the collection in the form of a tree itself.

Now, anyway so this tree is basically considered a new what is called the collection and which is
included in the another framework, the map framework. Other than this tree structure there is
another structure, it is called the table. So table is also one sort of form, it is called the map
framework. Now, here the question is arises that in what sense the two frameworks are different
from each other, collection framework and map framework?

In this regard, I just want to mention one thing, in the collection that we have learned the objects
are stored as a group of objects, they are stored either in array or in linked list or set, whatever it
is. Whereas in map framework objects are stored, but objects are stored with two what is called
the entities. One entity is basically the object itself that needs to be stored and another entity is
called key of the objects.

What is the concept of key, that I can explain you with a simple example? Consider books, there
is a number of books that needs to be stored in a programme suppose. So we can create a class
book and accordingly create several objects of the type book. So book is a, book class can
facilitate to create a collection of objects, they are called book objects.

Now fine, this is good idea, so this is idea so far collection framework maintains, but the map
framework enhance the concept little bit in a other sense to facilitate a faster retrieval, storing as
well as reading the data from the memory. So, faster storing and as well as faster retrieval. Now
how it is possible? The idea it is basically incorporating key to each object. What is actually key?

For a book object, for example, the accession number of the book is basically is a key of the
book. This means that if we give an input as an accession number then a corresponding book can
be retrieved or if we want to store a book object, we can store both the information that is the
accession number and the object itself.
Likewise, student record, for example, another example, student record has many attributes like
their name, their department, their date of joining, date of birth, marks obtained, different
courses, like, like, so a record can be of using many data you need. Now in order to have a
compact representation of the student, what we can use? We can use a number, let it be roll
number.

Roll number can be anything like, anyway so what I can say is that objects along with a key, that
means which is basically represents uniquely a particular objects can be used to store and
retrieve in a an efficient way. Now again let us come to the key, key is what? As I said key is an
identification number like accession number or roll number. A key value can be of any type
again, it can be integer, roll number say 1, 2, 3, 4 like, it can be a string like accession number
maybe A, then 2018 and B number say 4, 5.

So these are the strings, string is a basically alpha, is a collection of alpha numeric characters.
Even it can be any objects again. So a key can be anything. Now what is the basic difference
between the collection framework and map framework is that, while we are storing or retrieving
an element into a collection or from a collection, we, in collection framework we only consider
the object itself, object can be of any type.

Whereas, in case of map framework, we represents, we consider two entities, one is called the
key and another the object itself, where key and entity both can be of any type. So this is the
concept. Now based on this concept, the idea it is that how the map framework, Java developer
has proposed.
(Refer Slide Time: 8:57)

So in this lecture we will try to have a quick understanding about the different in the map
framework. Compare to collection framework, map framework is relatively simple, is a small, it
has only three interfaces namely map, then sorted map and navigable map. So three interfaces
are, in these three interfaces, the different methods are declared. So that all these methods, a Java
programmer can use in their Programme to manipulate their collections or we can say their maps.

Now as you know all the interfaces needs to be implemented, so for the implementation of all
those interfaces, Java supports in java.util several classes out of which abstract map is an abstract
class. And there are several real classes are there like enum map, hashmap, linked hashmap, tree
map and some less important, so far from the programmer’s utility point of view are called weak
hashmap and identity hashmap.

All these classes are inherited from the abstract class, abstract map. Abstract map implements
map interface. Now, again in the interface hierarchy, sorted map is basically is inherited from the
map and navigable map is basically inherited from the sorted map. And by virtue of inheritance
as you saw whatever the methods those are there in map interface are also available in their
sorted map or navigable map.

And in addition there are some of their own methods those are defined their interface. And again
I repeat all classes basically implements all the methods, so methods are declared in interface and
they are actually implemented in the class. Therefore, using all these class, you can create
objects, objects of that particular class. For example, using hashmap you can create an object of
hashmap.

Now an object of hashmap as you know, it is basically represents a collection and in this map
framework, a collection consist of two pairs values, one is called the key and another is the value
itself. So key and value, a group of keys and values constitute a collection in the map framework.
So this is basically the fundamental what is called a concept, so far the map framework is
concerned.

(Refer Slide Time: 11:50)

Now there are few interfaces we have already mentioned, like map, stored map and navigable
map. There is one inner class it is basically defined within the map interface itself; it is called the
map.Entry. Now, so these are the interface total that is there in the map framework and so far the
classes are concerned as we have already mentioned enum map, hashmap, tree map, linked
hashmap and weak hashmap and the identity hashmap are the classes those are defined there.
And all these interfaces and classes are defined there in java.util package where collection
framework is also defined there.
(Refer Slide Time: 12:37)

Now we will quickly have a, quick round about the different interfaces those are there. First of
all the map, it is basically the parent of interfaces belongs to this map framework and this map
frameworks basically facilitates many methods which basically useful to manipulate any map of
collection. Now entry is also is an inner class of this map, it also basically decides how an
element, that means an element in map framework you can remember, it is basically key value
pair.

Unlike in collection only the value or object, here key and object pair in a map. So this is
basically inner class so far map.Entry in concerned. Navigable map it is basically like map only
but here in case of map, suppose you want to search a particular value or objects and then to
search it you have to provide a key, if you do not provide key you will not be able to search, that
is one important concept.

Even if you want to insert, first you decide a unique key, if the key is not unique then it is again a
problem, whatever it be. So key has to be provided. Now in case of map, exactly if the key
matches with the stored elements those are there, then fine, you can retrieve it. If you do not find,
then navigable actually closest map, closest match, nearby this one it will say.

In some application it is required, exactly not the key value, but near about the key value, so that
you can finally filter it and then get the actual value if you want from the filter value. And sorted
map is same as map again, but the difference is that in case of map the elements that mean key
value pair are not necessarily to be stored in a order, in an order, but in case of sorted map, all the
objects are stored in the sorted order. Here sorted order is based on the key values.

(Refer Slide Time: 14:43)

So these are the different maps interfaces and now we will quickly check what are the different
methods are there in the map interface. Now map interface is a generic type, is basically
represents this collection T and T specifies the object of this one. Now different methods those
are there in the map framework I have listed in this tabular form that is for your future
consultation so that you can go through this table and then learn each method separately.

As in this is a short lectures period, so it is not possible to discuss each and every method but
here I just want to highlights about what exactly the methods about. All these methods basically
gives support, gives basically the facilities in order to how a particular key value can be created,
how key value and then object can be associated, how key value given a key value, how a
particular object can be searched, how given a key value and object how it can be inserted, how
given a key value and object how it can be removed and so many things are there.

Now for example here, in the method, that is the clear method. As the name implies we are
already familiar to the clear method what it does for us in the context of collection. Here also
clear method if we call for a collection that is a map collection, then it will clean all the entries
those are there in the collection. So it basically, if you want to remove all the entries under the
map collection then it will, you can use this method.

Now compute, here is basically the idea about how you can compute a key value for an object.
So this method sometimes required for a given object if you want to generate the key values like
this one. So there are different compute methods, compute present, compute absent, simple
compute that means if a key value is present then how a new key value for a given object can be
computed. If key value is absent, then what should be the methodology by which it can be
computed and so many things are there.

(Refer Slide Time: 16:54)

Now in addition to this there are few, some methods are there which basically to know the status
of the map collection, for example, the contents. Content means, should be particular object is
present in the collection or not. And then the get method, get method means given an object what
is the key value and vice versa, given a key value what is the object. And put value also it is
there, equals method is also there, for example how to find if the two objects are same or not that
belongs to this collection and likewise.

(Refer Slide Time: 17:32)

So there are many methods like this there are many methods to traverse the map collections also
merging the two collection into a single collection using merge method. And whether a
collection is empty or not, all these things can be obtained. If you want to remove a particular
elements form the map collection, you can use the remove method from there. Put method is
there, put conditionally, that means if a particular element is present in the collection already
then if you want to insert it, how it can be done.
(Refer Slide Time: 18:02)

So there are so many facilities that is given by the Java developer and you can just simply as a
fingertip use them in your Programme. Now regarding how to use all those methods, how to
create collection, whether it is a simple collection or map collection we definitely study in due
time and then you will be able to understand the strength of this java.util.

(Refer Slide Time: 18:24)

Now this is the map framework and a map interface and you know map framework is the parent
framework, parent interface and there is another is called a sorted map, which basically inherits
the map framework itself. And as I told you the sorted map is similar to the map only but only it
basically store the elements that is internally, you do not have to bother how to store. The Java
compiler will take the enough what is called the facilities, enough what is called the step to add if
you want to add to an existing collection a new object, then it will basically store them in a
sorted manner.
(Refer Slide Time: 19:13)

So basic idea of this sorted map interface is that faster retrieval, faster storing of the object
compared to the simple map, using the map methods. Likewise, the methods in map, in stored
map also there sorted map, there are many such methods available, they are, actually here it is a
mistake, it is not the stored map, it should be sorted map.

Anyway, and so they are basically same method in addition to all the methods those are there in
the map interface. Now some new methods like comparator that means if some objects needs to
be consider which is not of numeric type is a usually specially defined, custom defined user type
then it is your, it is the programmer’s responsibility to define how two objects can be compared.

So compare to method we will discuss about how the method can be redefined for your own
objects, so that means you have to, this is a method we can say that abstract method sort of thing
or the method it is there a default and you can overwrite this method according to your need in
your class. So, whenever you defined this class, you can create a compare to method which is an
overwrite.

So you can extend, if you create a new Programme, you can just extend sorted map or you can
extend hashmap or linked hashmap or T map, whatever it is there and then you redefine the
compare to class there. Now, there is again few methods are there, those are special to this sorted
map like first key as it represents that in the collection which is the starting key in the collection.
It returns the key object. Similarly, last key or just opposite to the first key and then it can tail
map. Tail map means, if you can give a key value then corresponding to this key value, what are
the next objects having key value after that one. So this key is a lower bound, then what is the
upper key values that is there, so starting from the start values of the key. So these are the few
special methods those are there in the sorted map interface.

(Refer Slide Time: 21:20)

Now let us come to the discussion of interface navigable map. It is just like a map only but the
idea is that here retrieval of entries based on the near about matches of the key values or the
closest match that is based on the search. This is the only thing that it is the different here, it
extends again the sorted map and declares the behavior of a map the supports for retrieval or
addition or deleting, whatever it is there, same methods are there.

(Refer Slide Time: 22:08)

So there are few methods which are there in navigable map of it owns, here we have mentioned
the methods. First of all ceiling entry, ceiling entry and floor entry are the two concepts that is
used that okay what will be the highest values, what is the lowest value (concept) like and there
is another method like descending, descending key set that means if you want to find in a
descending order of the key values for a given key, you can use it and there is a first entry
method, the floor entry method, the floor key method, these are the method basically related to
key, how we can find its closest values to it.

(Refer Slide Time: 22:46)

Then higher key, the lower key and then there is a navigable key, those are the methods. Higher
entry methods and then headmap method, all these methods are basically the different way that
the key and the value object pair can be managed, can be handled. So that is why the different
methods have been incorporated in the navigable map interface.
(Refer Slide Time: 23:11)

And then map dot entry is a inner class, it does not have much special about but only it consider
that how a key value pair can be enter into a map.
(Refer Slide Time: 23:20)

So it does not so much have its own much method but only few methods are there get key, get
value, hash code create, set value create and then whether the two values are equals or not, all
this things are there. So it is basically some peculiar reason of having a limited scope for the data
security, the map dot entry interface has been introduced.
(Refer Slide Time: 23:46)

Now let us come to the classes. There are few classes, we have already mentioned, apart from
this abstract map, the other important classes those are there is enum map, the hashmap, tree map
and linked hashmap. So mainly these are the (three) four, four classes are very crucial and useful
so far the map framework dealing is concerned. So as a programmer you have a very good
knowledge about all these four classes.

Now again I repeat all these four classes are basically the implementation of the interfaces that
we have just now discussed like map, that sorted map and then navigable map, whatever it is.
Now this means that all the methods that we have discussed so far interface is concerned, they
are implemented, they are in the classes, in addition some classes have their own methods also.
So if I do not mention any methods, which already we have discussed about interface, it implies
that these methods by virtue of implementation are there.

(Refer Slide Time: 24:59)

Now we will discuss about which are the different methods are there in the, in their own classes
actually.

(Refer Slide Time: 25:07)

So we will discuss about, let us start about the enum map. Now I just want to say enum is
basically is a kind of object. Like integer is a kind of object, float is a kind of object, enum is a
kind of object. This basically is the, enum is a type of object I can say. Now what is a type of
object? It is basically enumerated that means a particular value you can fix, just like enumerated
value. For example, an object say Sunday, another object is Monday, another object is Tuesday,
so these are the basically enumerated object.

I can say day is a enum type and in this enum type what are the different values that is possible?
Sunday, Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, like this, so seven what is
called objects. So enumerated is basically just like a set and it has only certain range of values
which is possible for the type. So enumerated is basically is a special type and it obviously all the
elements those are there in the range are all unique.

So in that case the enumerated is closely resemble with set type. Now here so map is concern,
each object should be associated with the key value. So Sunday is an object, its key value maybe
I can give it 0, Monday is another object I can give the key value 1, Tuesday is another object I
can give key value 3, so each key values is there. Now if we want to create a large collection of
type enum, then definitely for the fastest or efficient mechanism you should use both key values
as well as the enumerated values.

(Refer Slide Time: 26:49)

Now this way this enum map is basically it is there and as we can say that it basically extend
abstract map for use with enum keys.
(Refer Slide Time: 27:00)

So here key is basically the enumerated type and the object can be of any type like. Now, again
enum map is a class, so this means it has two elements, one is called the K, K extends enum, that
means K always should be of enumerated type. Beyond any enumerated type, it cannot be, that
mean K cannot be a number or it is string like. So this is the specialty so far this key values is
concerned because I told that key can be of any type.

This means that any type means number, or string or any user’s defined type but for this enum
map specialty, the key value should be of enumerated type. And then V is the object, that is
value is actually which you want store. There means a key value should be enumerated type by
which a particular objects can be associated and if we want to access a particular object, then we
should access through its enumerated key.
(Refer Slide Time: 27:57)

And it has only few methods, these constructors, only three constructors. The first constructor as
you see here the enum, this is a very simple constructor to create a collection, what we have to,
we have to give a input is that, what is the type of key values and key value itself. So by this way
an object can be created. And the second constructor, basically it gives both key values and then
the value itself. So key is always extends the map key and extends what is called the way. And
this is basically you know this a bound is there, upper bound.

So that means it is basically applicable map object only. And then, this is another, where we can
give the extension, it is the same way, this basically map K and it is basically is upper bound
extends V and it is a enumerated map. So these are the different way, we will see exactly with an
example, how the different objects can be created using this map. And yes one thing is that, the
enum map class, it does not have any exclusive method of its own.
(Refer Slide Time: 29:13)

Now next let us discuss about hashmap class. Hashmap class again is a child class of abstract
map, this means that whatever the methods those are declared there in map is also accessible to
hashmap in addition to its own methods.
(Refer Slide Time: 29:33)

Now let us see what are different way that hashmap collection can be created, as I already
mentioned that it is not a single pair, it is basically a single value. The hashmap should be key
and then the object value itself. So this is the composition that the collection should have.

(Refer Slide Time: 29:52)

Now the constructors, those are there in hashmap, they are pretty simple, it is just similar to just
this one. So, there hashmap this is default constructor that we can have. This is the default
constructor, we do not have to pass anything but sometimes you can pass many information to it.
For example, hashmap in capacity, so initially when you are to create a map collection, you can
mention what will be the size. It is for basically efficient memory management, if you say that
okay my collection have only size 100, you can mention it.

However, the initial capacity that you mention and if later on if you found that it require 1000
entries to be maintained then absolutely it is not an issue because the size of the map collection
automatically grow itself depending on the integer there. Now for this automatic or dynamic
progression of this map collection you can better for the memory management, if you know in
which increment that the memory should grow.

So in order to do these things, you can mention the fill ratio. For example, 75 percent, if the 75
percent is filled then you can go for increasing into another 10 size or these kind of thigs are
there. So default capacity is 16th but if you do not mention, for example different constructor
type you use then a map will be created with a default capacity as 16.

But if you can mention the flip ratio or say 0.5 or 0.8 or whatever it is there, it basically,
automatically grow these things. So these are the few constructors those are there in order to
create the map collection for your application. There are certain methods, for the hashmap also
there is no exclusive method other than the method those are there in map interface.

(Refer Slide Time: 31:54)

Now let us come to the tree map. It is basically same as the hashmap, only the difference, the
difference is from the internal point of view, not that programmer’s point of view. From the
internal point of view is that hashmap is basically stored using indexed representation, whereas
the tree map is used to store the data in linked representation.

Now there are three representation I told, one is the indexed, another is the sequential, another
linked representation. Array for example is an indexed representation, linked list for example is a
linked representation or sequential representation rather and tree structure is follows a linked
representation. What is a linked representation? That we shall understand when we will discuss
about the concept of tree that is there in the theory of computer science, data structure point of
view.

(Refer Slide Time: 32:52)

Anyway so tree map is basically the extension of the abstract class, it basically follows a tree
concept, a tree structure to store the key value pairs for the different objects.
(Refer Slide Time: 33:07)

So, two elements are to be there, so it is basically a collection by virtue of key objects and value
objects for each key value pairs of objects.

(Refer Slide Time: 33:18)

Now it also includes several constructors. The constructor, first one is the default constructor tree
map, the second is basically another constructor by which how the tree can be built, there
actually you have to define the comparator method because if you want to store, you just define
objects which has not the defined method of comparator.
Then you should define this comparator method, in that you just define type and then use it here.
And then tree map, it is basically the last constructor you can note. This basically tree map,
sorted map K extends Vm. So here also the tree can be store, tree collection can be stored using
the concept of sorted map that we have already discussed about the interface.

So that can be easier and again the tree map methods, if you see there are not any specific
method but those are the method specified by the navigable interface are there. So this means
that tree map methods includes all the methods those are there in the map interface as well as
those are there in navigable map interface is the part of this interface. So this is the tree map,
now next is basically linked hashmap class.

(Refer Slide Time: 34:47)

Linked hashmap pass use the linked representation, it is actually a sequential representation I can
say, sequential representation of the map collection object.
(Refer Slide Time: 34:57)

So here idea is that, it is basically allow insert, insertion order iteration. Actually the idea is that
if you see the sequential representation, if you want to add a new entry into a collection, it
always add in a particular order. So ordering is basically in what order the element will be
inserted. So here all the entries will be stored in an order but that order is not according to the
key values. This order is according to the order of insertion. So the first element which is
inserted should be at the head or the front and then the last element will be at the tail, at the end
actually, so it is layer. And in, so this way the elements are maintained in a sorted order.

(Refer Slide Time: 35:53)

There are few constructors belongs to this class. The constructors are defined here, the default,
first one is the default constructors. The second one is basically the constructor suppling key
values and object value, both together if you want to add more entities. Then the next, the third
constructor is basically creating a map collection specifying the size, default size is 16. And then
the last one is basically size and fill ratio both can be mentioned in order to create a new
collection of type linked hashmap.

(Refer Slide Time: 36:34)

Now there are certain classes also there. All the classes those are there in the map method are
basically defined for this linked hashmap method. It does not have any special methods of it
there.

(Refer Slide Time: 36:50)

And two other methods, two other classes like identity hashmap class and, sorry the first one is
weak hashmap class and the next one is identity hashmap class. These two classes are basically
for the use of Java developer not for the programmer. So it is basically at the system level
development, the Java developer defines classes to be used for maintaining the Java packages
and other, so we should not, as the programmer does not have any interest of it, so we should not
discuss in details about them.

(Refer Slide Time: 37:26)

Now finally I will come to here that whatever the map collections we have discussed, they
broadly can be categorized into four, one is hashmap which is of type indexed data structure is
follows, tree map is basically follow linked structure and then the linked hash structure is follow
basically sequential concept and the enum map is basically follow a string structure or bit string
form actually it is there.

(Refer Slide Time: 37:56)

Now regarding the discussion about all these collection concepts you can consult the link that is
given, the first link is basically useful link one because it is discuss in a very user friendly, reader
friendly manner. For advanced user the oracle tutorial document is very nice, you can go through
it. In the oracle tutorial document also you can find the discussion along with very small piece of
examples to elaborating many things are there. Third that there are many what is called the net is
also available in the net like geeks for geeks, Java T point or all these things, there also very nice
representation for easy to understand concept it is there. Thank you for your attention.