Number Systems

-Abhinav Singh (140101002)

-Rajan Garg (140101057)

Introduction
A numeral system (or system of numeration)
is a writing system for expressing numbers,
that is, a mathematical notation for
representing numbers of a given set, using
digits or other symbols in a consistent
manner.
It can be seen as the context that allows the
symbols "11" to be interpreted as the binary
symbol for three, the decimal symbol for
eleven, or a symbol for other numbers in
different bases.

Contents
1. Brief History
2. The Number Sense
3. Classification of Numbers
4. Number System(Definition)
5. Nth Number System
6. Decimal System
7. Binary System
8. Hexadecimal System
9. Does number system affect
rationality?
10.Conversion of Decimal to Binary
11.Motivation of the Algorithm
12.Proof of Correctness
13.C++ code Decimal to Nth conversion
algorithm
14.Conversion of Binary to Decimal
15.References

Brief History

Thousands of years ago there were no numbers to

represent two or three. Instead fingers, rocks, sticks or
eyes were used to represent numbers. There was little
need for a numeric system until groups of people formed
clans, villages and settlements and began a system of
bartering and trade that in turn created a demand for
currency. How would you distinguish between five and
fifty?
Paper and pencils were not available to transcribe
numbers. Other methods were invented for means of
communication and teaching of numerical systems.
Babylonians stamped numbers in clay by using a stick
and depressing it into the clay at different angles or
pressures and the Egyptians painted on pottery and cut
numbers into stone.

Evolution of Numbers

Quipu - An Inca Counting System

Imagine, if you will, a highly advanced civilization. This

civilization rules over a million or more people, they built
vast cities, developed extensive road systems. Now
imagine being able to do all this without a written
language. They did this using a memory tool made of
knotted strings called a quipu
Yet, it is unknown how a quipu was used but it is believed
that the strings were more or less used like a decimal
system.

The Mayan Number System

The Mayan number system dates back to the fourth
century and was approximately 1,000 years more
advanced than the Europeans of that time. This system is
unique to our current decimal system, which has a base
10, in that the Mayan's used a vigesimal system, which
had a base 20.

This system is believed to have been used because, since

the Mayan's lived in such a warm climate and there was
rarely a need to wear shoes, 20 was the total number of
fingers and toes, thus making the system workable.

The Greek Number System

The Greek
numbering
system was
uniquely
based upon
their
alphabet.
The original Greek alphabet consisted of 27 letters and
was written from the left to the right. These 27 letters
make up the main 27 symbols used in their numbering
system.
Problems with Ancient Number Systems1. In the above examples we see that to represent a
number we require a huge number of digits or
symbols. So to minimize the space required to
represent a number we need a more compact
numbering system. Space complexity is O(n). This
led to problems since not enough paper or rocks
were available.
2. No concept 0, neither the tally system nor the Roman
numeral system had the concept of 0.
3. No concept of bit addition. That is arithmetic
operations couldnt be done on individual digits

Sense of Numbers

The number sense is not the ability to count, but the ability to
recognize that something has changes in a small collection.

Some animal species are capable of this.

The number of young that the mother animal has, if changed, will be
noticed by all mammals and most birds. Mammals have more developed
brains and raise fewer young than other species, but take better care of
their young for a much longer period of time.
Many birds have a good number sense. If a nest contains four eggs, one
can safely be taken, but when two are removed the bird generally
deserts. The bird can distinguish two from three

One might think people would have a very good number sense, but as it
turns out, people do not. Experiments have shown that the average
person has a number sense that is around four.

Classification of Numbers

Number Systems
A number system is a writing system for
expressing numbers, that is, a mathematical
notation for representing numbers of a given set,
using digits or other symbols in a consistent
manner.
The value of a number akak-1a1a0 which is in
base x system is given by (where 0ai<x for
0xk)
akx k + ak-1x

k-1

++a1x 1+a0x

Ideally, a numeral system will:

Represent a useful set of numbers (e.g. all
integers, or rational numbers)
Give every number represented a unique
representation (or at least a standard
representation)

 Reflect the algebraic and arithmetic structure of

the numbers.

Few important points:

1.A number system with base X contain only
numbers from 0,1,2.X-1.
Example:
A binary system contains only 1 and 0s.
A octal system(base 8) contains only
0,1,2,3,4,5,6,7.
A hexadecimal system(base 16) contains only
0,1,2,3,4,5,6,7,8,9,A,B,C,D,E,F.
2.A number in one system can be changed to a
number
in another system.
Example:
A number 1310 is nothing but 11012 .
3.Value of a number akak-1a1a0 which is in
base x
system is given by. :
akxk + ak-1xk-1 + + a1x1 + a0x0.

Nth Number System

In general, if b is the base, one writes a number in
the numeral system of base b by expressing it in
the form
anbn + an 1bn 1 + an 2bn 2 + ... + a0b0
And writing the enumerated digits anan 1an
2 ... a0 in descending order.
The digits are natural numbers between 0 and b
1, inclusive.
That is, Digits={0,,b-1} if b-1<=9, otherwise
we might have to other symbols for example, in
case of Hexadecimal number system we have to
use Alphabets {A,B,C,D,E,F} too.
But total number of usable digits= b.
In general, numbers in the base b system are of
the form:

The numbers bk and bk are the weights of the

corresponding digits.

Decimal Number System

The decimal numeral system (also called base 10)
has ten as its base. It is the numerical base most
widely used by modern civilizations.
Base (also called radix) = 10
o 10 digits { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }
Digit Position
o Integer & fraction
Digit Weight
o Weight = (Base)

Position

Magnitude
o Sum of Digit x Weight
Formal Notation
o ak10 k + ak-110k-1 + + a1101 + a0100
Example:
(493)10 = 4*102 + 9*101 + 3*100.

Positional Notation

Position

...

1000

100

10

0.1

0.01

...

...

Weight

Digit

Decimal example weight

Decimal example digit

More about decimal system.

The Chinese counting rod system and the HinduArabic numeral system (the latter descended from
Brahmi numerals) are the first sources of decimal
system.
Ten fingers on two hands, the possible starting
point of the decimal counting.
Ten is the number which is the count of fingers and
thumbs on both hands (or toes on the feet). The
English word digit as well as its translation in many

languages is also the anatomical term for fingers

and toes. In English, decimal (decimus < Lat.)
means tenth, decimate means reduce by a tenth,
and denary (denarius < Lat.) means the unit of ten.

Binary Number System

Base (also called radix) = 2
o 2 digits { 0, 1 }
Positional notation number system
o Left most bit is most significant bit or
MSB
MSB

Binary Digit

LSB

28

27

26

25

24

23

22

21

20

256

128

64

32

16

o Right most bit is least significant bit or

LSB
Formal Notation
o ak2 k + ak-12k-1 + + a121 + a020
Example:
(101.01)2 = 1*22 + 0*21 + 1*20 + 0*2-1 +
1*2-2

Number of Binary Digits

(bits)

Common Name

Bit

Nibble

Byte

16

Word

Binary Number Names

Why is Binary System used in Digital?

On/Off signals provide very reliable operation.
Extremely low possibility of errors when
transferring data because it is easy to decide
the difference between the on and off state of
the circuit and some noise or disruption during
the transfer process will not disrupt that
decision.
Binary devices modelled after binary number
system are Simple and easy to build
Binary Numbers which are transmitted in the
form of binary signals are unambiguous (which
gives them noise immunity)
Flawless copy can be made of binary data

 Anything that can be represented with some

sort of pattern can be represented with pattern
of bits.
Since entire processing in computer happens in
binary digits (base 2), we have our memory
cards, pen drives memories as a powers of 2
i.e., 1GB, 2GB, 4GB, 8GB, 16GB, 32GB etc.

Hexadecimal Number System

Base = 16
o 16 digits { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C,
D, E, F }
Weights
o Weight = (Base) Position
Magnitude
o Sum of Digit x Weight
Formal Notation
o Example
(2E6.A3)16 = 2*162 + E*161 + 6*160 + A*16-1
+ 3*16-2
Hexadecimal has its own advantages:

Binary coded byte values can range from 0 to

255 (decimal) but may be more conveniently
represented as two hexadecimal digits in the
range 00 through FF. Hexadecimal is also
commonly used to represent computer
memory addresses.

Does number system affect

rationality?

Note that a number has a terminating or repeating

expansion if and only if it is rational; this does not
depend on the base. A number that terminates in
one base may repeat in another (thus 0.310 =
0.0100110011001...2).
An irrational number stays aperiodic (with an
infinite number of non-repeating digits) in all
integral bases.

Thus, for example in base 2, = 3.1415926...10 can

be written as the aperiodic
11.001001000011111...2.

Conversion from decimal to binary

number system
There are many ways to convert a decimal number
to a binary number. Few of them are listed below

Informal Way
1.Given a number N in decimal system.
2.Calculate the highest power of 2 which is less
than N. Let it be some K.

Eg:(Lets Solve the problem of converting N=3810

to binary) Highest power of 2 which is less than
38 is 32 (25).
3.Subtract the number K from N and output
1(i.e., push this into a stack).
Eg: When we subtract 32 from 38, we get 6 as
result. Push 1 into stack .
4.Now modify N =result obtained and K = next
least power of 2
Eg: The result obtained after subtraction is 6. So,
Now make N=6 and K=16(Next least power of 2
i.e., 25-1=24)
5. Now compare K with N.
a. If K is greater than N (k dont have capacity to
contribute to N), then output 0 (push 0 into
stack) and modify K =Next highest power of 2.
b. If K is less than N (k have capacity to
contribute to N), then output 1(push 1 into stack)
and subtract K from N . Modify N= result of
subtraction and K =Next highest power of 2.
Eg: K=16 and N=6. K>N (k dont have capacity
to contribute to N). Hence output 0 and K=8.
6. Repeat step v. until N=0 and then pop all
items from the stack.
Finally, 3810= 1001102

Eg: Convert 9710 into binary system.

Sol:
Let N=97. Highest power of 2 less than 97 is
64. Hence, k=64.
Compare k with N .k<N (subtractable) .Hence
[output 1] and subtract k from N
Modify N=N-k=97-64=33. (Here output means
either print or enqueue into a queue)
Modify k=Next highest power of 2 = 32.
Compare k with N .k<N (subtractable) .Hence
[output 1] and subtract k from N. Modify N = N-k
=33-32=1. Modify k=Next highest power of 2= 16.
Compare k with N. k>N (not subtractable).
Hence [output 0]. Modify k=Next highest power of
2= 8.
Initial value of N=97
Ste
Value Value
Modifie Modifie Outpu
p
of N
of K
d value d value t (In
No.
of N
of K
Queu
e)
1

97

64

33

32

3
4
5
6

01
01
01
01

16
08
04
02

9764=33
3332=01
01
01
01
01

32

16

08
04
02
01

0
0
0
0

01

01

0100
1
01=00
Since N=0, stop here and dequeue from queue.

Compare k with N. k>N (not subtractable).

Hence [output0]. Modify k=Next highest power of
2= 4.
Compare k with N. k>N (not subtractable). Hence
[output 0]. Modify k=Next highest power of 2= 2.
Compare k with N. k>N (not subtractable).
Hence [output 0]. Modify k=Next highest power of
2= 1.
Compare k with N .k<N (subtractable) .Hence
[output 1] and subtract k from N. Modify N = N-k
=1-1=0. Since, N =0, halt the process here.
Now dequeue the queue if you have used a
queue. So corresponding binary number is
11000012
Hence, 9710=11000012
DISADVANTAGES:
Note that, for larger numbers we need a table
containing still higher powers of 2.
For larger numbers it will be very difficult to
anticipate the number in terms of powers of 2

Procedural Way

Case I: For Natural Numbers

i. Let a no. N= akak-1ak-2a1a0 be a decimal
number to be converted into base x system.
ii. Divide N by x. Let quotient=q and output the
remainder (push into a stack)
iii.
Now assign N=q.
iv.
Repeat steps ii. And iii. Until N=0.
v. Pop out all the numbers from stack and print
them. Hence the binary number is displayed.

Eg : Lets convert 4310 into binary system.

Sol: Binary system have a base =2.
Divide 43 by 2 and collect remainder. Divide the
quotient again by 2 and collect remainder and
repeat the procedure until we get quotient as 0.

Direction
of
Counting

Note that the numbers written after hyphen in each

row is the remainder of the corresponding division
performed. Write the remainders in the direction of
arrow mark and finally answer is
4310 = 1010112

Eg : Lets convert 15610 into binary system.

Direction
of
Counting

15610 = 100111002
Eg: Lets convert 6710 into septal system.
Sol : Septal system have base =7.
Divide 67 by 7 and collect remainder. Divide the
quotient again by 7 and collect remainder and
repeat the procedure until we get quotient as 0.
Direction
of
Counting
Note that the numbers written after hyphen in each
row is the remainder of the corresponding division
performed. Write the remainders in the direction of
arrow mark and finally answer is
6710 = 1247

Eg : Lets convert 29310 into nonary (base 9)

system.
Sol: Nonary system have a base 9. So, instead of x,
put 9 and solve the problem. i.e., Divide 293 by 9
and collect remainder. Divide the quotient again by
9 and collect remainder and repeat the procedure
until we get quotient as 0.

Direction
of
Counting

Note that the numbers written after hyphen in each

row is the remainder of the corresponding division
performed. Write the remainders in the direction of
arrow mark and finally answer is
29310 = 3559

Case II: For Numbers with Decimal points

i. Convert the integer part of number N into base
x number system. This will be the integer part
of corresponding number of base x.
ii. Now we have to convert fraction part of N into
corresponding base x number.
iii.

Let fraction part on N be F = .a0a1a2ak

Multiply F with x. output the integer part of

product (either in a queue or in a array or
directly print it) and modify F=fractional part of
the product.
Eg: Let N = 34.2310 and x = 5 then F =
0.23
Multiply F with 5. Product =5 x 0.23 =1.15.
Output 1 and
modify F = 0.15
iv.
Repeat step iii. Until you get fractional part
as zero.
v. Dequeue the queue completely if you have
kept the output in a queue.

vi.
Sometimes for some numbers we may not
be able to get the fractional part=0, even after
so many steps. In that case, we get a precision
error. This is because, every fractional part
cannot be expressed as a sum of 0.5, 0.25,
0.125 and so on.

Eg : Convert 0.687510 into binary system.

Sol : Since the integer part of the number is zero,
the corresponding binary number is zero.
Now, lets convert the fractional part of the
number(0.6875) into corresponding binary number.

Direction of
Counting of
integer parts

Since the fractional part became zero in 4th step,

we can halt the multiplication here. On counting
integer parts in shown direction, we get
corresponding binary number as 0.1011.
Hence, 0.687510=0.10112

Eg: Convert 1404.25410 into hexadecimal

system.
Sol: In hexadecimal (base 16) system, we have
totally 16 symbols for representing numbers. Those
are 0,1,2,3,4,5,6,7,8,9,A,B,C,D,E,F .
For example,
Integer part and fractional part must be converted
to hexadecimal system separately. So, first lets
convert 1404 into hexadecimal system.

Direction
of
Counting

Hence, 140410=57C16

Now, lets convert fractional part (0.254) into

hexadecimal system.
0.254 x 16 = 4.064 [integer part = 4]
0.064 x 16 = 1.024 [integer part = 1]
0.024 x 16 = 0.384 [integer part = 0]
0.384 x 16 = 6.144 [integer part = 6]
0.144 x 16 = 2.304 [integer part = 2]
0.304 x 16 = 4.864 [integer part = 4]
0.864 x 16 = D.824 [integer part = D]
0.824 x 16 = D.184 [integer part = D]
Since we are not achieving fractional part to be
zero even after so many steps, we can leave the
problem here. But note that there will be some
precision error.
Hence, 0.25410 = 0.410624DD16
Hence, 1404.25410 = 57C.410624DD16

Motivation for the Algorithm

Motivation for the Technique without
point
Let given number N=akak-1ak-2a1a0 in decimal
system .Let the number be brbr-1br-2b1b0 in base x
system.
Now our task is to find br,br-1,br-2,b1,b0 .
We know that , N= brxr + br-1xr-1 + br-2xr-2 + +
b1x1 + b0
Since every number of br,br-1,br-2,b1,b0 are present
in number system with base x, they are all less
than x.
i. Divide N by x. Remainder is b0 and quotient is
brxr-1 + br-1xr-2 + br-2xr-3 + + b1 . So, we found
b0 (nothing but our remainder) .So, output the
found remainder. Now let N= brxr-1 + br-1xr-2 +
br-2xr-3 + + b1 .
Clearly we can get all digits by consequent
divisions
ii. Again divide N by x. Remainder is b1 and
quotient is brxr-2 + br-1xr-3 + br-2xr-4 + + b1 .

So, we found b1 (nothing but our remainder)

.So, output the found remainder.
iii.
Repeat above steps until we got
quotient=0 (that is, until every value of br,br1,br-2,b1,b0 is found).

Motivation for the Technique with

point.
Now lets see for conversion of fractional part of
decimal number to corresponding fractional part of
base x system.
Let fractional part of given number N (base
10) , is F = .a1a2a3ak .Let the corresponding
fractional part in base x system be R
= .b1b2b3br .
Now, our task is to find R which is nothing but
to obtain the values of b1, b2, b3,br. The value
of this R in decimal is given by b1x-1 + b2x-2 +
b3x-3 ++brx-r.
Hence, .a1a2a3ak = b1x-1 + b2x-2 + b3x-3 +
+brx-r
i. Multiplying by x on both sides yields x(.a1a2a3ak
)= b1 + b2x-1 + b3x-2 ++brx-r+1
Since, every one of b1, b2, b3, br are less than x ,
Fractional part of the product = b2x-1 + b3x-2 +
+brx-r+1
Integer part of the product = b1.

 Hence b1 is found and output the value of b1

(nothing but integer part of the product)
ii. Now fractional part is b2x-1 + b3x-2 ++brx-r+1 .
Let this be F . Multiply F with x which results in
Fractional part of the product = b3x-1 ++brx-r+2
Integer part of the product = b2.
Hence b2 is found. Output b2 (which is nothing but
integer part of the product)
iii. Repeat the step i. and ii. Until we get all values
of b1, b2, b3,br. Sometimes value of r may be large
and we may not be able to find all the values of b1,
b2, b3,br .In that case , we get precision error.

Proof for the Algorithm (Decimal to

Binary)
For Conversion of Integral Part
Let us prove by Induction that our claim is right
for all numbers greater than 2.
Base caseConversion of
(0)10=(0)2
(1)10=(1)2
(2)10=(10)2
which are trivially right
Inductive StepTo prove that binary conversion of number n is
right assuming binary conversion of all numbers
less than n is right.
Now from the 1st step in the algorithm we take
remainder of n divided by 2 and then we push the
answer in the stack and continue the process with
number = n/2. Using Mathematical Principle of
Strong Induction the Binary Conversion of n/2
from the algorithm is right.
If n is evenNow since if n is even then this algorithm writes
the final answer as (an an-1an-2a1) a0 (a0 =0).

20a0+21a1+22a2.2nan=
0+2*(20a1+21a2+22a32n-1an) = 0+2*(n/2)
If n is even this algorithm gives the right result.
If n is oddIf n is odd then the algorithm will write the final
answer (an an-1an-2a1)a0 (a1=1).
20a0+21a1+22a2.2nan =
1+2*(20a1+21a2+22a32n-1an) = 1+2*(n-1)/2= n
(since n is odd n/2 will evaluate to (n-1)/2 by the
algorithm)
If n is odd this algorithm gives the right result.
QED.

Proof of Conversion of Fractional PartFact 1The algorithm tries to find a binary representation
that is the closest to the fractional part and at the
same time is less than equal to the number. In
other words the number outputted by the
algorithm will not in any scenario be greater than
the number whose conversion we are trying to
find
We want to prove that the kth digit after decimal
printed by the algorithm is correct.

Proof by Induction
Here O refers to the original fractional number.
And Fk refers to the fractional number used by the
algorithm in the k step. For example In Binary
conversion of 0.54 O=0.54
And the next number that is used is 2*0.541=0.08=F2 since it used in the 2nd step.
Base case1st digit after decimal is correct and the F2= 2*O a-1.
O=a-1/2 +a-2/4 + a-3/8 .
a-2/4 + a-3/8.<= 1/4+1/8+1/16.=1/2
(Follows from the sum of an infinite GP)
if O>=1/2 the a-1 will be 1 ,else 0.
O*2>=1 then a-1 is 1 else 0 . (1) (This is what
the algorithm does)
If a-1 is 1 then clearly 2*O>=1 and F2=2*O-1
(assignment)
Else F2=2*O (assignment)

Hence F2 is 2*O - a-1 where we have calculated a-1

correctly.

Inductive Step
Using induction we know that the algorithm at k-1
step uses
Fk-1 =2k-1O - 2k-2a-2 - 2k-3a-3 .-a-k+1
(2)
Using Original Equation
O=a-1/2 +a-2/4 + a-3/8. + a-k+1/2k-1 + a-k/2k + a-kk+1

1/2
Doing some manipulations we can reduce it to
2k-1O - 2k-2a-2 - 2k-3a-3 .-a-k+1 =a-k/2 +a-k-1/4
Using (2) we get
Fk-1=a-k/2 +a-k-1/4
This is equivalent to the base case and we can
use the same logic to prove that the algorithm
computes the k digit after decimal correctly and
also that
Fk =2*Fk-1 - a-k.
Fk = 2kO - 2k-1a-2 - 2k-4a-3 .-2*a-k+1-a-k
This proves the inductive step

QED

C++ code for the Algorithm to convert

decimal to Nth number system
#include<iostream>
#include<stdlib.h>
using namespace std;
class stack //Declaring stack class
{
private:
int top,num;
int *a;
public:
stack(int x=10) //to set the default value of size as 10
{
a=new int[x];
top=-1;
num=x;
}
void push(int data)

{
if(top<=num-1)
{
a[top++]=data;
}
else if(top==num-1)
{
num=num+5;
a=(int *)realloc(a,num*4);
a[top++]=data;
}
}
int pop()
{
if(top==-1)
cout<<"stack is empty"<<endl;
else
return a[--top];
}
int gettotal()

{
return (top+1);
}
};
int main()
{
int n,x,i;
float k,data,f;
stack a;
cout<<"enter a number : ";
cin>>data;
cout<<endl<<"Enter the destination number system:";
cin>>x;
n=(int)data; //storing integer part of data in n
f=data-n; //storing fractional part of data in f
while(n!=0)
{
a.push(n%x); //pushing the remainder
n=n/x; //updating the value of n
}

cout<<endl<<data<<" in base "<<x<<" system is : ";

int num=a.gettotal();
for(i=0;i<num;i++) //popping all bits
{
cout<<a.pop();
}
cout<<"."; //for giving a decimal point7
for(i=0;i<20;i++) //to get a precision upto 20 digits
{
k=f*x;
cout<<(int)k;
f=k-(int)k;
if(f==0)
break;
}
}

ALGORITHM TO CONVERT BINARY

TO DECIMAL

To convert a binary number into a decimal

number, we multiply each bit to it corresponding
weight. The weights are in the sequence of 20, 21,
23,, 2n,
So the decimal representation is nothing but the
sum of the product of each bit with its
corresponding weight.
n

(anan-1an-2...a0.a-1a-2a-3...) 2 = a
0

2 +

a -k 2-k
1

Eg(101.11)2 = 1*22+0*21+1*20+1*2-1+1*2-2
= (5.75)10
Eg- Convert 1010110102 from binary to
decimal.

2 + 8 + 16 + 64 + 256 = 34610

Other base system to Decimal System

 Step 1 - Determine the column (positional)

value of each digit (this depends on the
position of the digit and the base of the
number system).
Step 2 - Multiply the obtained column values
(in Step 1) by the digits in the corresponding
columns.
Step 3 - Sum the products calculated in Step
2. The total is the equivalent value in decimal.
Eg:
Step

Octal Number

Decimal Number

Step 1

258

((2 x 81) + (5 x 80))10

Step 2
Step 3

258
258

(16 + 5 )10
2110

Octal Number: 258 = Decimal Number: 2110

References
https://en.wikipedia.org/wiki/Numeral_syst
em
http://www.math.wichita.edu/history/topics
/num-sys.html
http://betterexplained.com/articles/numbe
rs-and-bases/
http://www.tutorialspoint.com/computer_fu
ndamentals/computer_number_conversion
.htm