• Programming in C

• Recursion
• Arrays
• Stacks
• Queues
• Linked lists
• Trees
• Binary search trees
• Binary heaps
• Graphs
• Limits
• Continuity and differentiability
• Maxima and minima
• Mean value theorem
• Theorems of integral calculus
• Evaluations of definite and improper integrals
• Taylor series
• Partial derivatives
• Multiple integrals
• Vector identities
• Directional derivatives
• Vector space, basis
• Linear dependence and independence
• Matrix algebra, rank, determinants
• System of linear equations
• Eigenvalues and eigenvectors
• LU decomposition
• Random variables
• Uniform, normal, exponential, Poisson and binomial distributions
• Mean, median, mode and standard deviation
• Conditional probability and Bayes theorem
• Correlation and regression analysis
• Propositional and first order logic
• Sets, relations, functions
• Partial orders and lattices, groups
• Graphs: connectivity, matching, coloring
• Combinatorics: counting, recurrence relations, generating functions.

https://www.youtube.com/watch?v=b-UZJVdLbXc&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZkAz
DETERMINANTS

• For a 𝑛 × 𝑛 determinant, the max number of terms during expansion will be 𝒏!

• For a diagonal matrix, upper triangle matrix and lower triangle matrix, the determinant is the
product of the terms in the principal diagonal. Also, if any term of the principal diagonal is
zero, then the determinant will also be zero.
• A matrix has determinant as zero if the matrix has –
o One row has all elements as zero
o One row is proportional to another i.e. one row when multiplied by an integer, we get
another row
• If we interchange two rows/columns of a matrix, then the sign of the determinant changes.

Let us assume a 2𝑛 × 2𝑛 matrix in the following format –

Then, for this matrix (say A), the determinant is given by –

Δ = (1 − 𝑎2 )𝑛

1
det(𝐴) = det(𝐴𝑇 ) =
det(𝐴−1 )

For any matrix 3 × 3 or above, if the numbers (in row major format) are consecutive, then the
determinant will be zero. For example –
11 12 13
det (14 15 16) = 0
17 18 19
TYPES OF MATRICES

• Row – Of the form 1 × 𝑛

• Column – Of the form 𝑛 × 1
• Square – Of the form 𝑛 × 𝑛
• Rectangular – Of the form 𝑚 × 𝑛
• Singular – Whose determinant is zero
• Null – Each entry is zero
• Diagonal – Square matrix where 𝑎𝑖𝑗 = 0 ∀ 𝑖 ≠ 𝑗
• Upper triangle – Square matrix where 𝑎𝑖𝑗 = 0 ∀ 𝑖 > 𝑗
• Lower triangle – Square matrix where 𝑎𝑖𝑗 = 0 ∀ 𝑖 < 𝑗
• Identity – Diagonal matrix with principal diagonal elements being 1
• Scalar – Diagonal matrix with principal diagonal elements being some scalar 𝐾
• Symmetric – Matrices where 𝐴𝑇 = 𝐴
• Skew-Symmetric – Matrices where 𝐴𝑇 = −𝐴. The principal diagonal elements of a skew-
symmetric matrix is always 0.
• Orthogonal – Matrices where 𝐴𝑇 = 𝐴−1
• Unitary – Matrices where 𝐴𝜃 = 𝐴−1
• Idempotent – Matrices where 𝐴2 = 𝐴
• Involuntary – Matrices where 𝐴2 = 𝐼
• Hermitian – Matrices where 𝐴𝜃 = 𝐴
• Skew-Hermitian – Matrices where 𝐴𝜃 = −𝐴

NOTE

When we take the complex conjugate of every element of a matrix, then the resulting matrix is called
the conjugate matrix and is denoted by 𝑨∗ . Additionally, we can define the trajugate of a matrix as
follows –

𝐴𝜃 = (𝐴∗ )𝑇 = (𝐴𝑇 )∗

NOTE

det(𝐴 + 𝐵) ≠ det(𝐴) + det(𝐵)

INVERSE OF 3X3 MATRICES

https://byjus.com/maths/inverse-of-3-by-3-matrix/

(𝐴𝐵)−1 = 𝐵 −1 𝐴−1
For a 𝑚 × 𝑛 matrix, the number of minors of size 𝑟 × 𝑟 is given by –
𝑚
𝑁𝑜 𝑜𝑓 𝑚𝑖𝑛𝑜𝑟𝑠 (𝑜𝑓 𝑠𝑖𝑧𝑒 𝑟) = 𝑟𝐶 ∗ 𝑛𝑟𝐶

RANK AND TRACE

The number of linearly independent rows of a matrix is termed as the rank of that matrix. Rank is also
the highest order of minors of a matrix that are not all singular.

• The NULL Matrix is the only matrix with rank (denoted by 𝜌) as zero.
• Elementary row/column ops don’t change the rank
• Rank of a scalar matrix is the same as the order of the matrix
• For a matrix of order 𝑚 × 𝑛, the max rank can be 𝑎 where 𝑎 = min(𝑚, 𝑛)
• For a 𝑛 × 𝑛 non-singular matrix, 𝜌(𝐴) = 𝜌(𝐴−1 ) = 𝜌(𝐴𝑇 ) = 𝜌(𝑎𝑑𝑗 𝐴) = 𝑛
• If a matrix has all the elements as the same value, then the rank is 1

NULLITY

It is the number of zero rows in the row echelon form. The sum of rank and nullity is equal to the
number of rows of the matrix.

SYSTEM OF LINEAR EQUATIONS

If a system of linear equations have no solution, then they are said to be inconsistent. Otherwise, if
they have either a unique solution of infinitely many solutions, they are said to be consistent. Also,
when the constant term of each equation is zero, then the system is said to be homogeneous. Else, it
is called non-homogeneous.

Solution to non-homogeneous equations

We write the equations in the form of the matrices –

𝐴𝑋 = 𝐵

• Write the augmented matrix which is of the form [𝐴 ∶ 𝐵]

• Convert the augmented matrix to a row echelon form
• Get the ranks 𝜌(𝐴) and 𝜌(𝐴 ∶ 𝐵)
o If the ranks are not equal, then the equations are inconsistent with no solution
o If the ranks are equal and the same as the number of equations (rows), then there is
a single solution (trivial)
o If the ranks are equal but lesser than the number of equations, then there are
infinitely many solutions (non-trivial).

Solution to homogeneous equations

Homogeneous equations always result in a consistent solution. This is because the 𝐵 matrix is always
0 so 𝜌(𝐴) = 𝜌(𝐴 ∶ 𝐵). Now, if rank of A is the same as the number of equations, then there will be a
single solution. If the rank of A is lesser than the number of equations, then there will be infinitely
many solutions.

NOTE

Let us assume we have 2 vectors – 𝑋 and 𝑌. Vectors are either row or column matrices. Now, if we
have –

𝑋𝑌 𝑇 = 0 𝑋𝑋 𝑇 ≠ 0 𝑌𝑌 𝑇 ≠ 0
Then the two vectors are said to be mutually orthogonal vectors. On the other hand, if we have –

𝑋𝑌 𝑇 = 0 𝑋𝑋 𝑇 = 1 𝑌𝑌 𝑇 = 1
Then the two vectors are said to be mutually orthonormal vectors.

EIGENVALUES AND EIGENVECTORS

For a matrix 𝐴, first we solve the equation –

|𝐴 − 𝜆𝐼| = 0

This gives us values of 𝜆 which will be our Eigenvalues. Now, if we solve the set of homogeneous
equations –
[𝐴 − 𝜆𝐼][𝑋] = 𝑂
Then, the solution vectors 𝑋 are called Eigenvectors. Basically, each eigenvalue has its own
eigenvector.

NOTE

If a matrix has eigenvector 𝑋, then the vector 𝑐𝑋 may also be an eigenvector where 𝑐 is some constant.
Also, the solution to the equation [𝐴 − 𝜆𝐼][𝑋] = 𝑂 is always non-trivial.

Question

Calculate the Eigenvalues and Eigenvectors for the matrix –

5 4
𝐴=[ ]
1 2
Answer

𝜆 = 1,6
−1 4
𝑋1 = 𝑐 [ ] 𝑋2 = 𝑐 [ ]
1 1

NOTE

• Sum of eigenvalues = Sum of principle diagonal elements = Trace of a matrix

• Product of eigenvalues = determinant of the matrix
• For a diagonal matrix or an upper triangular matrix, the diagonal elements are the eigenvalues
LOWER – UPPER (LU) DECOMPOSITION

This says that every matrix can be expressed as a product of a lower triangle and an upper triangle
matrix.
𝑎11 𝑎12 𝑎13 1 0 0 𝑢11 𝑢12 𝑢13
𝐴 = [𝑎21 𝑎22 𝑎23 ] = 𝐿𝑈 = [𝑙21 1 0] [ 0 𝑢22 𝑢23 ]
𝑎31 𝑎32 𝑎33 𝑙31 𝑙32 1 0 0 𝑢33

CAYLEY – HAMILTON THEOREM

It states that every square matrix satisfies its own characteristic equation. Basically, the equation we
get when we solve 𝐴 − 𝜆𝐼 is the characteristic equation. If we substitute 𝐴 instead of 𝜆, then we get
NULL matrix. This is the statement of the theorem. For eg, let us take –
1 2
𝐴=[ ]
0 2
Here, the characteristic equation will be –

|𝐴 − 𝜆𝐼| = 𝜆2 − 3𝜆 + 2

Substituting 𝐴 instead of 𝜆, we get –

1 6 3 6 2 0 0 0
𝐴2 − 3𝐴 + 2𝐼 = [ ]−[ ]+[ ]=[ ]=𝑂
0 4 0 6 0 2 0 0
Hence, the theorem is proven

EXPRESSING HIGHER ORDER MATRICES

Here is an interesting concept. It states that every order of a matrix can be represented as a function
of its original form. Basically, it states that for a matrix 𝐴 –

𝐴𝐾 = 𝑐1 𝐴 + 𝑐2 𝐼
Now, we just need to calculate the constants. This can be done using eigenvalues. Suppose the matrix
𝐴 has eigenvalue 𝜆. Then,

𝜆𝐾 = 𝑐1 𝜆 + 𝑐2
1 2
For eg, if we have 𝐴 = [ ], then what will be the value of 𝐴9 ? For 𝐴, we have –
0 2
𝜆 = 1,2
Thus, we get –

19 = 𝑐1 1 + 𝑐2

29 = 𝑐1 2 + 𝑐2
Solving the 2 equations, we get –

𝑐1 = 511 𝑐2 = −510
Hence, we get –
 𝐴9 = 511𝐴 − 510𝐼

VECTOR LINEAR DEPENDENCE vs INDEPENDENCE

For a set of vectors, write them in a matrix and then calculate the rank of the matrix. If the matrix rank
is the same as the number of rows, then the vectors are linearly independent. Otherwise, they are
linearly dependent.

Question

Answer

First, we write the vectors in the form of a matrix as follows –

1 2 3
𝐴 = [−2 −1 −5]
𝜆 5 7𝜆
If the vectors are linearly dependent, then the rank of the matrix should be less than 3 and hence the
determinant of the matrix should be zero. Hence, we get –
5
det(𝐴) = 0 => 𝜆 =
14

VECTOR SPACE & SUBSPACE

It is a set of vectors that share a certain common property. A subset of the vector space is called a
vector subspace.

LINEAR SPAN
In the above case, let us write –

𝐿(𝑋) = 𝑋1 + 2𝑋2 + 3𝑋3 + ⋯ + 𝑁𝑋𝑁

This is a valid linear span. Now, suppose we have 2 vector spaces 𝑉 and 𝑋 and let’s say that 𝑉 spans 𝑋.
Then we can write –

𝑉𝑖 = 𝑐1 𝑋1 + 𝑐2 𝑋2 + 𝑐3 𝑋3 + ⋯ + 𝑐𝑛 𝑋𝑛
Watch -
https://www.youtube.com/watch?v=SzMQzgQHIkg&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZkAz&inde
x=12

59:00 onwards

Basically, if we need to find whether a vector space spans another vector space, we create a set of
non-homogeneous equations. If these set of equations have either a unique or infinitely many
solutions, then the vector space is spanned. Else, it is not spanned

BASIS OF VECTORS

Suppose we have 2 vector spaces –

𝑆⃗ = {𝑖̂, 𝑗̂, 𝑘̂ }

𝐴⃗ = {3𝑖̂ + 𝑗̂ + 2𝑘̂ , 𝑖̂ + 𝑗̂ − 𝑘̂ , 7𝑖̂ − 14𝑗̂ − 21𝑘̂ }

In this case, we can use 𝑆⃗ to get 𝐴⃗ by altering the weights. Thus, we can say that 𝑺
⃗⃗ is the basis of 𝑨
⃗⃗ .
For a vector space 𝑆⃗ to be a basis of vector space 𝐴⃗, we need –

• All vectors in 𝑆⃗ should be linearly independent

• 𝐴⃗ should be spanned by 𝑆⃗

Question
Answer

We can see that the 3 vectors are linearly independent (their matrix has rank 3). Now for spanning,
we form the augmented matrix (non-homogeneous equations) as follows –
1 0 0∶ 𝑥
[0 1 0 ∶ 𝑦]
0 0 1∶ 𝑧
We can see that 𝜌(𝐴) = 𝜌(𝐴𝐵) = 3. Hence, there exists a unique solution. Hence, 𝑩 spans 𝑽. Since
both the conditions satisfy, 𝑩 is the basis vector of 𝑽

DIMENSION

The number of distinct elements in the basis of the vector space is the dimension of the vector space.
It has a few properties –

• dim(𝑉1 ∪ 𝑉2) = dim(𝑉1) + dim(𝑉2) − dim(𝑉1 ∩ 𝑉2)

• Dimension of null space is nullity
• The dimension is the number of linearly independent vectors in that space.

https://www.youtube.com/watch?v=W8uG3SUGjVM&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZkAz&in
dex=14

The above video is for practice questions.

LIMITS

In calculus, we express numbers in 3 forms –

• Determinant
• Indeterminant
• Undefined

To check the form, first equate the number expression to a variable (say 𝑥) and then find the values of
𝑥. If –

• 𝑥 takes a unique value, then the number is in determinate form

• 𝑥 takes infinitely many values, then the number is in indeterminate form
• No value of 𝑥 exists, then it is an undefined form. These are usually represented by ±∞

For example, let us take the number 15⁄3. We can write –

15
=𝑥
3
 𝑥=5
Since 𝑥 takes a unique value, the number is in determinate form. Now let us take the example –
15
=𝑥
0
0. 𝑥 = 15
We can see that no value of 𝑥 can satisfy the equation. Hence, the number is in undefined form. Finally,
let us take the example –
0
=𝑥
0
0. 𝑥 = 0
Here, 𝑥 can take infinitely many values and hence the number is in indeterminate form. Given below
are some standard indeterminate forms –

Each of the 7 forms above can eventually be expressed as 𝟎⁄𝟎

CALCULATING LIMIT OF A FUNCTION

To calculate the limit, we need to first calculate both – the Left-Hand Limit (LHL) and the Right-Hand
Limit (RHL). Suppose we take the function –
sin 𝑥
𝑓(𝑥) =
𝑥
Now, we get –

𝐿𝐻𝐿 = lim− 𝑓(𝑥) = 1− (𝑎 𝑙𝑖𝑡𝑡𝑙𝑒 𝑙𝑒𝑠𝑠 𝑡ℎ𝑎𝑛 1)

𝑥→0

𝑅𝐻𝐿 = lim+ 𝑓(𝑥) = 1− (𝑎 𝑙𝑖𝑡𝑡𝑙𝑒 𝑙𝑒𝑠𝑠 𝑡ℎ𝑎𝑛 1)

𝑥→0

We see that –

𝐿𝐻𝐿 = 𝑅𝐻𝐿
When this happens, we declare that the limit exists at that point and the value of the limit is 1.

Question

Calculate the limit at point 𝑥 = 0 for the following function –

2𝑥
𝑓(𝑥) =
𝑥
Answer

We can re-write the function as follows –

2 𝑥≠0
𝑓(𝑥) = {0 }
𝑥=0
0
Hence, we get –

𝑓(0) = 𝑢𝑛𝑑𝑒𝑓𝑖𝑛𝑒𝑑
But, let’s calculate the limits –

𝐿𝐻𝐿 = lim− 𝑓(𝑥) = 2

𝑥→0

𝑅𝐻𝐿 = lim+ 𝑓(𝑥) = 2

𝑥→0

Since 𝐿𝐻𝐿 = 𝑅𝐻𝐿, we can say that the limit exists and the value is 2.

Question

Find the limit –

lim [𝑥] 𝑤ℎ𝑒𝑟𝑒 [. ] 𝑖𝑠 𝑡ℎ𝑒 𝑔𝑟𝑒𝑎𝑡𝑒𝑠𝑡 𝑖𝑛𝑡 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛

𝑥→0

Answer

𝐿𝐻𝐿 = lim−[𝑥] = −1
𝑥→0

𝑅𝐻𝐿 = lim+[𝑥] = 0
𝑥→0

Since the LHL and RHL don’t match, the limit doesn’t exists at 𝑥 = 0
STANDARD FORMS AND APPROACHES

MODEL 1 – L’HOSPITAL’S RULE

If a limit is of the form 0⁄0, then we can use this rule –

𝑓(𝑎) 𝑓′(𝑎) 𝑓′′(𝑎) 𝑓′′′(𝑎)

lim = lim = lim = lim =⋯
𝑔(𝑎) 𝑔′(𝑎) 𝑔′′(𝑎) 𝑔′′′(𝑎)
Question

Find the limit

sin 𝑥
lim
𝑥→0 𝑥

Answer

This is a 0⁄0 format. Hence, we can use the L’Hospital rule –

sin 𝑥 cos 𝑥
lim = lim =1
𝑥→0 𝑥 𝑥→0 1

Question

Find the limit

sin 𝑥
lim [ ]
𝑥→0 𝑥
Answer

We can also use the series expansion to solve this problem.

𝑥3 𝑥5 𝑥7
sin 𝑥 = 𝑥 − + − + ⋯
3! 5! 7!
Hence,
 sin 𝑥 𝑥2 𝑥4 𝑥6
= 1 − + − + ⋯ = 1−
𝑥 3! 5! 7!
Therefore,
sin 𝑥
lim [ ]=0
𝑥→0 𝑥

FEW IMP OBSERVATIONS

sin 𝑥
lim = 1−
𝑥→0 𝑥

𝑠𝑖𝑛 −1 (𝑥)
lim = 1+
𝑥→0 𝑥
tan 𝑥
lim = 1+
𝑥→0 𝑥

𝑡𝑎𝑛−1 𝑥
lim = 1−
𝑥→0 𝑥

MODEL 2

The problem format will be in the form –

lim 𝑓(𝑥) 𝑔(𝑥) = 1∞

𝑥→𝑎

The solution in this case will be of the form 𝑒 𝑙 where –

𝑙 = lim [𝑓(𝑥) − 1]𝑔(𝑥)

𝑥→𝑎

Question

Find the limit

1
lim (1 + 𝑥)𝑥
𝑥→0

Answer

This is of the form 1∞ . Hence, we first find –

𝑙 = lim [𝑓(𝑥) − 1]𝑔(𝑥)

𝑥→𝑎
1
Here, 𝑓(𝑥) = 1 + 𝑥 ; 𝑔(𝑥) = 𝑥. Therefore, we get –

1
𝑙 = lim 𝑥 ∗ =1
𝑥→0 𝑥
Hence, we get –
𝟏
𝐥𝐢𝐦(𝟏 + 𝒙)𝒙 = 𝒆
𝒙→𝟎
MODEL 3

The problem format will be in the form –

lim 𝑓(𝑥) 𝑔(𝑥) = 00 𝑜𝑟 ∞0

𝑥→0

To get the solution, just take natural log on either side.

Question

Find the limit.

−𝑥
lim (1 + 𝑥 2 )𝑒
𝑥→∞

Answer

We take –
−𝑥
𝑦 = lim (1 + 𝑥 2 )𝑒
𝑥→∞

Taking natural log on both sides, we get –

−𝑥 2)
ln(1 + 𝑥 2 )
ln 𝑦 = lim 𝑒 ∗ ln(1 + 𝑥 = lim
𝑥→∞ 𝑥→∞ 𝑒𝑥
Now, we can apply the L’Hospital rule on the RHS

ln(1 + 𝑥 2 )
lim =0
𝑥→∞ 𝑒𝑥
Therefore, we get –

ln 𝑦 = 0
𝒆−𝒙
𝒚 = 𝐥𝐢𝐦 (𝟏 + 𝒙𝟐 ) =𝟏
𝒙→∞

MODEL 4

This is the final model and the limit is of the form ∞ − ∞. To solve this case, just try to rationalize and
simplify the expression so that it becomes one of the 3 other models – preferably Model 1.

Question

Find the limit.

1 1
lim ( 2
− )
𝑥→0 𝑥 𝑠𝑖𝑛2 𝑥
Answer

We can write the limit as –

 𝑠𝑖𝑛2 𝑥 − 𝑥 2
lim ( )
𝑥→0 𝑥 2 𝑠𝑖𝑛2 𝑥

Now, the limit is in the 0⁄0 form which is Model 1. We can now solve using L’Hospital rule and get the
answer.

LEIBNITZ RULE

COMMON SERIES EXPANSION

CONTINUITY

A function is said to be continuous around a point if –

𝐿𝐻𝐿 = 𝑅𝐻𝐿 = 𝐹𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢𝑒 𝑎𝑡 𝑡ℎ𝑎𝑡 𝑝𝑜𝑖𝑛𝑡 ≠ ∞

Based on the above definition, there are three types of discontinuities –

• Type 1 of Removable type

• Type 1 of Non-removable type
• Type 2
Type 1 Removable Discontinuity

This is the case where –

𝐿𝐻𝐿 = 𝑅𝐻𝐿 ≠ 𝐹𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢𝑒

Also note that all three values are finite. For example,
sin 𝑥
𝑓(𝑥) =
𝑥
At 𝑥 = 0, we get –

𝐿𝐻𝐿 = 𝑅𝐻𝐿 = 1− ≠ 𝑓(0)

This is called a removable type of discontinuity and can be removed by a simple re-definition of the
function as follows –
sin 𝑥⁄ 𝑥 ≠ 0}
𝑓(𝑥) = { 𝑥
1 𝑥=0
Now, the function is continuous around 𝑥 = 0

Type 1 Non-removable Discontinuity

This is the case where –

𝐿𝐻𝐿 ≠ 𝑅𝐻𝐿
For example, let us take the signum function itself –
1 𝑥>0
𝑠𝑔𝑛(𝑥) = { 0 𝑥 = 0}
−1 𝑥<0
Now, at 𝑥 = 0, we get –

𝐿𝐻𝐿 = −1 ; 𝑅𝐻𝐿 = 1
This cannot be made into a continuous function and hence is termed as a non-removable discontinuity.

Type 2 Discontinuity

This is the case where either LHL, RHL or the function value either tends to ±∞ or is oscillatory (can’t
know the exact value).

Question

Discuss the continuity of –

1
𝑓(𝑥) = 𝑎𝑡 𝑥 = 0
𝑥
Answer

Here,

𝐿𝐻𝐿 = −∞ ; 𝑅𝐻𝐿 = +∞ ; 𝑓(𝑥) = ±∞

Hence, the function has Type 2 discontinuity.

NOTE

Suppose we have a rational function as follows –

𝑓(𝑥)
𝐹=
𝑔(𝑥)
If the roots of 𝑔(𝑥) = 0 are real, then at those points, the function will tend to ∞ and hence will
become discontinuous. However, if the roots are imaginary, then the function will remain continuous
throughout the real plane.

DIFFERENTIABILITY

To calculate whether a function is differentiable around a point 𝑥 = 𝑎, we calculate the left hand and
right hand differential –
𝑓(𝑎) − 𝑓(𝑎 − ℎ)
𝐿𝐻𝐷 = lim
ℎ→0 ℎ
𝑓(𝑎 + ℎ) − 𝑓(𝑎)
𝑅𝐻𝐷 = lim
ℎ→0 ℎ
If, 𝐿𝐻𝐷 = 𝑅𝐻𝐷 = 𝑓𝑖𝑛𝑖𝑡𝑒 𝑣𝑎𝑙𝑢𝑒, then the function is differentiable at 𝑥 = 𝑎

Question

Answer

For a mod function, the point where the function becomes zero is the point of non-differentiability.
Let us take 𝑥 = −2
|−2 + 2| − |−2 − ℎ + 2| −|−ℎ|
𝐿𝐻𝐷 = lim = lim = −1
ℎ→0 ℎ ℎ→0 ℎ
|−2 + ℎ + 2| − |−2 + 2| |ℎ|
𝑅𝐻𝐷 = lim = lim =1
ℎ→0 ℎ ℎ→0 ℎ

As we can see, 𝐿𝐻𝐷 ≠ 𝑅𝐻𝐷 at 𝑥 = −2. That is the point of non-diff. However, the range asked is from
[0,4]. Therefore, there are no points of non-diff in the given range.
NOTE

Continuity is a necessary condition for differentiability. If a function is differentiable, then it is

continuous. Also, if a function is discontinuous, then it is non-differentiable.

MAXIMA AND MINIMA

A function can have 3 types of nature –

• Increasing
• Decreasing
• Constant

NATURE OF FUNCTION CONDITION

Strictly/Monotonically Increasing 𝑓(𝑎 + ℎ) > 𝑓(𝑎) > 𝑓(𝑎 − ℎ)
Increasing/Non-decreasing 𝑓(𝑎 + ℎ) ≥ 𝑓(𝑎) ≥ 𝑓(𝑎 − ℎ)
Strictly/Monotonically Decreasing 𝑓(𝑎 + ℎ) < 𝑓(𝑎) < 𝑓(𝑎 − ℎ)
Decreasing/Non-increasing 𝑓(𝑎 + ℎ) ≤ 𝑓(𝑎) ≤ 𝑓(𝑎 − ℎ)

The points where the function attains the max and min values are called maxima and minima
respectively. The points where the function doesn’t attain a max or a min value is called a saddle point.
To get the range of values where the function is either decreasing or increasing, we first find the
derivative of the function and plot the same. The range where the derivate is positive is the one where
the function is monotonically increasing and the range where the derivative is negative is the one
where the function is monotonically decreasing.

Question

Answer

First, we find the domain of the function. Here, we get –

𝐷𝑜𝑚𝑎𝑖𝑛 𝑜𝑓 𝑓(𝑥) = 𝑅
Now, we find the derivate of the function as follows –

𝑓 ′ (𝑥) = 2 ∗ (𝑥 2 − 4) ∗ 2𝑥 = 4𝑥 3 − 16𝑥
For Monotonically Increasing

4𝑥 3 − 16𝑥 > 0
Thus,

𝒙 ∈ (−𝟐, 𝟎) ∪ (𝟐, ∞)
For Monotonically Decreasing

4𝑥 3 − 16𝑥 < 0
Thus,

𝒙 ∈ (−∞, −𝟐) ∪ (𝟎, 𝟐)

Now, we can define the terms –

• Local maxima – The max value around a part of the graph

• Local Minima – The min value around a part of the graph
• Global maxima – The max value of the entire function
• Global minima – The min value of the entire function

To find the local maxima and minima, we need to examine the critical points which are –

• Points where 𝑓 ′ (𝑥) = 0

• Points of discontinuity
• Points of non-differentiability

1ST DERIVATIVE TEST

As per this test, point 𝑥 = 𝑎 is a local maxima if –

𝑓 ′ (𝑎 − ℎ) > 0 ; 𝑓 ′ (𝑎) = 0 ; 𝑓 ′ (𝑎 + ℎ) < 0

On the other hand, the point is a local minima if –

𝑓 ′ (𝑎 − ℎ) < 0 ; 𝑓 ′ (𝑎) = 0 ; 𝑓 ′ (𝑎 + ℎ) > 0

Question

Find the local maxima and minima for the function –

𝑓(𝑥) = (𝑥 2 − 4)2
Answer

First we need to find critical points. Since the function is both continuous and differentiable, we need
only look at the points where the derivate becomes zero.

𝑓 ′ (𝑥) = 4𝑥(𝑥 2 − 4) = 4𝑥(𝑥 + 2)(𝑥 − 2)

The critical points are 𝑥 = 0, +2, −2

For 𝒙 = 𝟎

𝑓 ′ (0 − ℎ) = 𝑓 ′ (−ℎ) = −4ℎ(2 − ℎ)(−2 − ℎ) = 4ℎ(ℎ2 − 4)

Where ℎ is a small positive value. We can see that –

𝑓 ′ (0 − ℎ) > 0
Similarly, we can see that –

𝑓 ′ (0 + ℎ) < 0
Therefore, we can conclude from the first derivate test that 𝒙 = 𝟎 is a local maxima.

Similarly, we perform the analysis for the other two critical points and find that the points 𝒙 = ±𝟐 are
local minima.

2ND DERIVATIVE TEST

First, we find the critical points. Then we calculate the second derivative at those points.

• If 𝑓 ′′ (𝑐) > 0, then the point is a local minima

• If 𝑓 ′′ (𝑐) < 0, then the point is a local maxima
• If 𝑓 ′′ (𝑐) = 0, then we need to differentiate further. The point 𝑥 = 𝑐 is a saddle point

LOCAL MAXIMA AND MINIMA OF 2-VARIABLE FUNCTIONS

Suppose we have a function –

𝑍(𝑥, 𝑦) = 𝑎𝑥 + 𝑏𝑦 + 𝑐
To find the local maxima and minima of the function, we first find the partial derivates wrt to 𝑥, 𝑦 –
𝜕𝑍
𝑝= =𝑎
𝜕𝑥
𝜕𝑍
𝑞= =𝑏
𝜕𝑦
𝜕2𝑍
𝑟= =0
𝜕𝑥 2
𝜕2𝑍
𝑠= =0
𝜕𝑥𝜕𝑦
𝜕2𝑍
𝑡= =0
𝜕𝑦 2
Equating both 𝑝 and 𝑞 to zero will give us the critical points of the form (𝑥, 𝑦). Once we have the critical
point, we calculate the value of –

𝐴 = 𝑟𝑡 − 𝑠 2

• If 𝐴 > 0 and 𝑟 < 0, then the point is the local max

• If 𝐴 > 0 and 𝑟 > 0, then the point is the local min
• If 𝐴 < 0, then the point is the saddle point
• If 𝐴 = 0, then we need to investigate further
Question

Answer

Let’s calculate the partial derivates –

𝜕𝑓
𝑝= = 2𝑥 + 6
𝜕𝑥
𝜕𝑓
𝑞= = 2𝑦
𝜕𝑦
𝜕2𝑓
𝑟= =2
𝜕𝑥 2
𝜕2𝑓
𝑠= =0
𝜕𝑥𝜕𝑦
𝜕2𝑓
𝑡= =2
𝜕𝑦 2
Equating 𝑝 = 0 and 𝑞 = 0, we get –

𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 𝑝𝑜𝑖𝑛𝑡 → (𝑥, 𝑦) = (−3,0)

We can also see that –

𝐴 = 𝑟𝑡 − 𝑠 2 = 4 > 0
Since 𝐴 > 0 and 𝑟 > 0, the point is the local minima.

INTEGRAL CALCULUS

When we do integration without limits, it is called an indefinite integral and when we do with limits,
it is called a definite integral. There are some properties of definite integrals –
NOTE
𝜋⁄
2
(𝑚 − 1)(𝑚 − 3)(𝑚 − 5) …
∫ 𝑠𝑖𝑛𝑚 (𝑥)𝑑𝑥 = { }∗𝐾
𝑚(𝑚 − 2)(𝑚 − 4) …
0

Where,
1 𝑚 𝑖𝑠 𝑜𝑑𝑑
𝐾 = {𝜋 𝑚 𝑖𝑠 𝑒𝑣𝑒𝑛}
⁄2

The same formula applies when 𝑠𝑖𝑛 is replaced by 𝑐𝑜𝑠.

NOTE

We can evaluate the GIF integral by splitting the limits into integral subsets as follows –
2 −2 −1 0 1 2

∫[𝑥]𝑑𝑥 = ∫ −3𝑑𝑥 + ∫ −2𝑑𝑥 + ∫ −1𝑑𝑥 + ∫ 0𝑑𝑥 + ∫ 1𝑑𝑥 = −𝟓

−3 −3 −2 −1 0 1
Also, we can have the case –
2 2 2

∫[𝑥 + 3]𝑑𝑥 = ∫[𝑥]𝑑𝑥 + ∫ 3𝑑𝑥 = 𝟏𝟎

−3 −3 −3

AREA BOUND BETWEEN CURVES

For example, let us try to find the area enclosed between the curves –

𝐶1 ∶ 𝑦 2 = 4𝑥

𝐶2 ∶ 𝑥 2 = 4𝑦
We can plot the graph as follows –

Thus, to find the area enclosed, we subtract the integrals as follows –

4
𝑥 2 𝟏𝟔
𝐴 = ∫ 2√𝑥 − = 𝒔𝒒 𝒖𝒏𝒊𝒕𝒔
4 𝟑
0

SPHERICAL COORDINATES
Here,

𝜃 ∈ [0, 𝜋]
𝜙 ∈ [0,2𝜋]
To convert the rectangular coordinates into spherical coordinates, we use –

𝑥 = 𝑟 sin 𝜃 cos 𝜙 𝑦 = 𝑟 sin 𝜃 sin 𝜙 𝑧 = 𝑟 cos 𝜃 𝑑𝑥𝑑𝑦𝑑𝑧 = 𝑟 sin 𝜃 𝑑𝑟𝑑𝜃𝑑𝜙

CUMULATIVE DISTRIBUTION FUNCTION (CDF)

If we have a random variable 𝑋, then the CDF of the same can be given as –

𝐹(𝑥) = 𝑃(𝑋 ≤ 𝑥) = 1 − 𝑃(𝑋 > 𝑥)

As a result, we get certain properties –

• 𝐹(∞) = 𝑃(𝑋 ≤ ∞) = 1
• 𝐹(−∞) = 𝑃(𝑋 ≤ −∞) = 0
• 0 ≤ 𝐹(𝑥) ≤ 1
• CDF is right continuous, i.e. 𝐹(𝑎) = 𝐹(𝑎+ )

PROBABILITY DENSITY FUNCTION (PDF)

It is the differentiation of the CDF –

𝑑𝐹(𝑥)
𝑓(𝑥) =
𝑑𝑥
As a result, we get certain properties –

• 𝑓≥0
∞
• ∫−∞ 𝑓(𝑥)𝑑𝑥 = 1
∞
• 𝑓(𝑥) will always be an even function. If it were an odd function, then the ∫−∞ limit would have
resulted in zero instead of 1.

Question

Answer

We can write that –

∞
1 −𝑥 𝟏
𝑃(𝑋 > 5) = ∫ 𝑒 ⁄5 𝑑𝑥 = ≈ 𝟎. 𝟑𝟕
5 𝒆
5
EXPECTATION, VARIANCE & SD

This is the mean or the average value of a random variable –

∞

𝐸[𝑋] = ∫ 𝑥𝑓(𝑥)𝑑𝑥 (𝑓𝑜𝑟 𝑐𝑜𝑛𝑡𝑖𝑛𝑢𝑜𝑢𝑠 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠) = ∑ 𝑥𝑖 𝑃(𝑋 = 𝑥𝑖 ) (𝑓𝑜𝑟 𝑑𝑖𝑠𝑐𝑟𝑒𝑡𝑒 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠)

−∞

Variance can be defined as –

𝜎 2 = 𝐸[𝑋 2 ] − (𝐸[𝑋])2
Standard Deviation can be defined as –

𝜎 = √𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒

Question

A fair coin is tossed thrice. Let 𝑋 be the number of heads. Find the mean, variance and SD of the same.

Answer

𝑋 = (0,1,2,3)
We can write –

𝑿 0 1 2 3
𝑷(𝑿) 1/8 3/8 3/8 1/8

As we can see, this is a discrete R.V. Hence, we can write –

3 3 3
𝐸[𝑋] = 0 + + + = 𝟏. 𝟓
8 4 8
3 3 9
𝐸[𝑋 2 ] = 0 + + + = 𝟑
8 2 8
𝜎 2 = 3 − 1.52 = 𝟎. 𝟕𝟓

𝜎 = √0.75 = 𝟎. 𝟖𝟔𝟔

IMP SERIES EXPANSIONS

DISTRIBUTIONS

• Uniform
• Triangular
• Gaussian/normal
• Rayleigh
• Exponential
• Binomial
• Poisson

Uniform Distribution

DISTRIBUTION FUNCTION MEAN VARIANCE

𝑓(𝑥)
1
𝑎 ≤ 𝑥 ≤ 𝑏} 𝑏+𝑎 (𝑏 − 𝑎)2
Uniform = {𝑏 − 𝑎
0 𝑒𝑙𝑠𝑒 2 12

𝑎+𝑚+𝑏 𝑎2 + 𝑚2 + 𝑏 2 − 𝑎𝑏 − 𝑏𝑚 − 𝑎𝑚
Triangular See image
3 18
Gaussian/ 1 −(𝑥−𝜇)2
𝑓(𝑥) = 𝑒 2𝜎2 𝜇 𝜎2
Normal 𝜎√2𝜋
𝑓(𝑥)
2
𝑥 −𝑥2 𝜋 4−𝜋
Rayleigh = {𝜎 2 𝑒 2𝜎 𝑥 ≤ 0} 𝜎√ 𝜎2 ∗
2 2
0 𝑒𝑙𝑠𝑒
𝑛 𝑟 𝑛−𝑟
Binomial 𝑟𝐶 𝑝 𝑞 𝑛𝑝 𝑛𝑝𝑞
−𝜆 𝑟
𝑒 𝜆
Poisson 𝜆 𝜆
𝑟!
−𝜆𝑥 1 1
Exponential 𝑓(𝑥) = { 𝜆𝑒 𝑥 > 0}
0 𝑒𝑙𝑠𝑒 𝜆 𝜆2

Question
Calculate the integral below.
∞
2
∫ 𝑒 −𝑥 𝑑𝑥
0

Answer

This is an interesting case. First thing to observe here is that the function is an even function –
2 2
𝑓(−𝑥) = 𝑒 −(−𝑥) = 𝑒 −𝑥 = 𝑓(𝑥)
Thus, we can say –
0 ∞
−𝑥 2 2
∫𝑒 𝑑𝑥 = ∫ 𝑒 −𝑥 𝑑𝑥
−∞ 0

Thus, our integral can be written as –

∞
1 2
𝐼 = ∫ 𝑒 −𝑥 𝑑𝑥
2
−∞

Now, it is just a matter of re-writing the expression –

2 1 −(𝑥−0)2
𝑒 −𝑥 = √𝜋 ∗ 𝑒 1
√𝜋 ∗ 1
This resembles a gaussian distribution with 𝜇 = 0 and 2𝜎 2 = 1. We can now use the property of a
gaussian distribution 𝐺(𝑥) –
∞

∫ 𝐺(𝑥)𝑑𝑥 = 1
−∞

Therefore, we get –

√𝜋
𝐼= ∗ 1 = 𝟎. 𝟖𝟖𝟔
2

CORRELATION AND COVARIANCE

𝐶𝑜𝑟𝑟𝑒𝑙𝑎𝑡𝑖𝑜𝑛 (𝑅𝑋𝑌 ) = 𝐸[𝑋𝑌]

𝐶𝑜𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝐶𝑜𝑣(𝑋, 𝑌) = 𝐸[(𝑋 − 𝜇𝑋 )(𝑌 − 𝜇𝑌 )] = 𝑅𝑋𝑌 − 𝜇𝑋 𝜇𝑌
Some tips –

• If 𝑋 and 𝑌 are orthogonal, then 𝑅𝑋𝑌 = 0

• If 𝑋 and 𝑌 are uncorrelated, then 𝐶𝑜𝑣(𝑋, 𝑌) = 0. As a result, 𝑅𝑋𝑌 = 𝜇𝑋 𝜇𝑌 or 𝐸[𝑋𝑌] =
𝐸[𝑋]𝐸[𝑌]
• If 𝑋 and 𝑌 are independent, then 𝐸[𝑋 𝑎 𝑌 𝑏 ] = 𝐸[𝑋 𝑎 ]𝐸[𝑌 𝑏 ]. This also means that if the R.V. are
independent, then they are also uncorrelated (𝑎 = 𝑏 = 1).
PROPERTIES OF VARIANCE

STATISTICS

Data can either be grouped or un-grouped. We can conclude that –

NOTE
For a data collection, we can write –
𝑀𝑜𝑑𝑒 = 3 ∗ 𝑀𝑒𝑑𝑖𝑎𝑛 − 2 ∗ 𝑀𝑒𝑎𝑛
If the distribution is symmetric, then we can write –
𝑀𝑒𝑎𝑛 = 𝑀𝑒𝑑𝑖𝑎𝑛 = 𝑀𝑜𝑑𝑒
If the distribution is not symmetric, then the distribution is said to be skewed.
𝑃𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑆𝑘𝑒𝑤 → 𝑀𝑒𝑎𝑛 > 𝑀𝑒𝑑𝑖𝑎𝑛 > 𝑀𝑜𝑑𝑒
𝑁𝑒𝑔𝑎𝑡𝑖𝑣𝑒 𝑆𝑘𝑒𝑤 → 𝑀𝑒𝑎𝑛 < 𝑀𝑒𝑑𝑖𝑎𝑛 < 𝑀𝑜𝑑𝑒

STANDARD DEVIATION
We can define –

∑(𝑥𝑖 − 𝑥̅ )2 𝑓𝑖
𝜎=√
∑ 𝑓𝑖

We can use SD to also define –

𝑀𝑒𝑎𝑛 − 𝑀𝑜𝑑𝑒
𝑆𝑘𝑒𝑤𝑛𝑒𝑠𝑠 𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 =
𝜎

Question
Find all the central tendencies.
Answer
Here,
𝑥𝑖 = {3,9,15,21,27,33,39}
Thus, we can get –
∑ 𝑥𝑖 𝑓𝑖 (3 ∗ 6) + (9 ∗ 11) + (15 ∗ 25) + (21 ∗ 35) + (27 ∗ 18) + (33 ∗ 12) + (39 ∗ 6)
𝑥̅ = =
∑ 𝑓𝑖 6 + 11 + 25 + 35 + 18 + 12 + 6
̅ = 𝟐𝟎. 𝟕𝟑
𝒙
For Median, we need to find the Median Class. The Median class is the class with the middle element
which is the 18 − 24 class. Thus, we can define –
𝐿 = 18 ; 𝐾 = 6 ; 𝑓 = 35

𝐹 = ∑ 𝑓𝑖 = 113 ; 𝐶 = 6 + 11 + 25 = 42

Thus, we can write –

𝐹
−𝐶
𝑀 = 𝐿 + (2 ) 𝐾 = 𝟐𝟎. 𝟒𝟖𝟔
𝑓

Now, we can calculate the Mode. To do so, we need to find the Modal Class which is the class with
max frequency. This is also 18-24. Thus, we get –
𝑙 = 18 ; 𝐹 = 35 ; 𝐹1 = 18 ; 𝐹−1 = 25 ; 𝐾 = 6
Thus, we can write –
𝐹 − 𝐹−1
𝑀𝑜𝑑𝑒 = 𝑙 + ( ) 𝐾 = 𝟐𝟎. 𝟐𝟐
2𝐹 − 𝐹−1 − 𝐹1

Question
Answer
Option C is the correct answer

PROPOSITIONAL LOGIC
This is a bunch of declarative statements that can either be “TRUE” or “FALSE”. To define the
statements, we use a bunch of connectives –
SYMBOL EXPRESSION DEFINITION
OR 𝐴⋁𝐵 This is equivalent to logical OR
AND 𝐴⋀𝐵 This is equivalent to logical AND
NEGATION ¬𝐴 This is equivalent to logical NOT
IMPLICATION 𝐴→𝐵 This means “if A then B”. It is equivalent to 𝐴̅ + 𝐵
IFF 𝐴⇔𝐵 This is equivalent with XNOR.
 • Tautologies – When the statements are always true.
• Contradictions – When the statements are always false.
• Contingencies – When the statements are sometimes true, sometimes false.
If two statements have the same truth table or are true under IFF statement, then they are said to be
Propositional Equivalent. Suppose we have an implication 𝑋 → 𝑌, then we say that 𝑋 is the hypothesis
and 𝑌 is the conclusion. For an implication, we have the following definitions –

• Inverse – If 𝑋 → 𝑌, then the inverse is written as ¬𝑋 → ¬𝑌

• Converse – If 𝑋 → 𝑌, then the converse is written as 𝑌 → 𝑋
• Contra-positive – If 𝑋 → 𝑌, then the contra-positive is written as ¬𝑌 → ¬𝑋
We can also define Normal Forms as follows –
• Conjunctive Normal Form – This is a series of ORs joined by ANDs. For eg - (𝑎 ∨ 𝑏) ∧ (𝑐 ∨ 𝑑)
• Disjunctive Normal Form – This is a series of ANDs joined by ORs. For eg - (𝑎 ∧ 𝑏) ∨ (𝑐 ∧ 𝑑)

PREDICATE LOGIC
A predicate is an expression of one of more variables defined on some domain. We can quantify it
using two main quantifiers –
• Universal Quantifier – It is the for all (∀) operator.
• Existential Quantifier – It is the there exists some (∃) operator.
Now, we can also define the quantifiers as follows –
∀𝑥𝑃(𝑥) = 𝑃(𝑥1 ) ∧ 𝑃(𝑥2 ) ∧ … ∧ 𝑃(𝑥𝑛 )
∃𝑥𝑃(𝑥) = 𝑃(𝑥1 ) ∨ 𝑃(𝑥2 ) ∨ … ∨ 𝑃(𝑥𝑛 )
Therefore, we can define their negations as follows –

¬∀𝑥𝑃(𝑥) = ¬𝑃(𝑥1 ) ∨ ¬𝑃(𝑥2 ) ∨ … ∨ ¬𝑃(𝑥𝑛 ) = ∃𝑥(¬𝑃(𝑥))

¬∃𝑥𝑃(𝑥) = ¬𝑃(𝑥1 ) ∧ ¬𝑃(𝑥2 ) ∧ … ∧ ¬𝑃(𝑥𝑛 ) = ∀𝑥(¬𝑃(𝑥))

RULES OF INFERRENCE
As the name suggests, we can use these to infer statements. The standard form is as follows –
𝑋
𝑌
𝑍
∴𝑆
The statements on the top (𝑋, 𝑌, 𝑍) are called hypothesis or premises. The statement at the bottom
(𝑆) is called the conclusion. Basically, we say that if premises 𝑋, 𝑌, 𝑍 are true, then the conclusion 𝑆 is
a tautology. Let us take an example –
𝑃→𝑄
𝑄→𝑅
∴𝑃→𝑅
How do we check this? Let us draw the truth tables –
 𝑷 𝑸 𝑹 𝑷→𝑸 𝑸→𝑹 𝑷→𝑹
0 0 0 1 1 1
0 0 1 1 1 1
0 1 0 1 0 1
0 1 1 1 1 1
1 0 0 0 1 0
1 0 1 0 1 1
1 1 0 1 0 0
1 1 1 1 1 1

We can see that whenever the hypothesis 𝑃 → 𝑄 and 𝑄 → 𝑅 are true, then the conclusion 𝑃 → 𝑅 is
also true making it a tautology. We can also understand this with intuition. Let us take –
𝑃 − 𝑌𝑜𝑢 ℎ𝑎𝑣𝑒 𝑡ℎ𝑒 𝑝𝑎𝑠𝑠𝑤𝑜𝑟𝑑
𝑄 − 𝑌𝑜𝑢 𝑐𝑎𝑛 𝑙𝑜𝑔𝑖𝑛 𝑡𝑜 𝑦𝑜𝑢𝑟 𝑒𝑚𝑎𝑖𝑙 𝑎𝑐𝑐𝑜𝑢𝑛𝑡
𝑅 − 𝑌𝑜𝑢 𝑐𝑎𝑛 𝑠𝑒𝑛𝑑 𝑎𝑛 𝑒𝑚𝑎𝑖𝑙
So this problem becomes – “If you have the password, you can login to your email account. If you can
login to your email account, you can send an email. Therefore, if you have the password, you can send
an email.”
There are some standard rules of inference. You need not remember this by-heart as you can prove
them. Just try to remember the “Modus” rules.

SET THEORY
We already know a bunch of things about sets, but there are some concepts about relations and
functions that need to be explored. First, let us explore the concept of Relations.
A relation 𝑥𝑅𝑦 between two sets 𝑥 and 𝑦 is basically a subset of 𝑥 × 𝑦. Therefore, if the cardinality of
the sets are |𝑥| = 𝑚 ; |𝑦| = 𝑛, then the cardinality of 𝑅 can vary from 0 ≤ |𝑅| ≤ 𝑚𝑛. For any relation
𝑥𝑅𝑦, the 𝑥 and 𝑦 values make up the domain and range respectively.
Types of Relations
TYPE DEFINITION
Null 𝑥𝑅𝑦 = 𝜙
Full 𝑥𝑅𝑦 = 𝑋 × 𝑌
Reflexive If (𝑎, 𝑎) is a part of the relation ∀𝑎 ∈ 𝐴
Identity 𝑅 = {(𝑥, 𝑥) | 𝑥 ∈ 𝑋}
′
Inverse 𝑅 = {(𝑏, 𝑎)| (𝑎, 𝑏) ∈ 𝑅}
Symmetric If 𝑥𝑅𝑦 implies 𝑦𝑅𝑥 ∀𝑥, 𝑦 ∈ 𝐴
Anti-symmetric If 𝑥𝑅𝑦 and 𝑦𝑅𝑥 implies 𝑥 = 𝑦, ∀𝑥, 𝑦 ∈ 𝐴
Transitive If 𝑥𝑅𝑦 𝑎𝑛𝑑 𝑦𝑅𝑧 implies 𝑥𝑅𝑧 ∀𝑥, 𝑦, 𝑧 ∈ 𝐴
Equivalence A relation that is reflexive, symmetric and transitive

FUNCTIONS
A function 𝑓: 𝑋 → 𝑌 is a relationship from elements in one set 𝑋 (also called domain) to another set 𝑌
(also called codomain). Basically, 𝑥 ∈ 𝑋 and 𝑦 ∈ 𝑌, if there is a unique (𝑥, 𝑦) ∈ 𝑅, then it is a function.
Here, 𝑥 is called the pre-image and 𝑦 is the image. These definitions are fine and all, but not really
important. The important stuff is the TyPeS oF fUnCtIoNs –

Injective Functions
These are one-to-one functions. A function 𝑓: 𝐴 → 𝐵 is said to be injective if for each value in 𝐵, there
exists only one pre-image in 𝐴. For example, 𝑓: 𝑁 → 𝑁, 𝑓(𝑥) = 𝑥 2 is injective as for each natural
number there is a unique square. On the other hand, 𝑓: 𝑅 → 𝑅, 𝑓(𝑥) = 𝑥 2 is NOT injective as real
numbers include positive and negative numbers and both give the same square. So, for one element
in co-domain, there are two pre-images in domain.
Surjective Functions
These are onto functions. If for a function, every element in the co-domain has at least one pre-image
in domain, then it is an onto function.
Bijective Functions
If a function is both injective and surjective, then it is bijective. For a function 𝑓: 𝐴 → 𝐵 to be bijective,
|𝑨| = |𝑩|.

NOTE

If 𝑓 and 𝑔 are injective, then 𝑓𝑜𝑔(𝑥) = 𝑓(𝑔(𝑥)) is also injective. Similarly, if 𝑓 and 𝑔 are surjective,
then 𝑓𝑜𝑔(𝑥) = 𝑓(𝑔(𝑥)) is also surjective.

GROUP THEORY
Let us consider a non-void set 𝐺 and a binary operator ⨁ such that for 𝑎, 𝑏 ∈ 𝐺, we get 𝑐 = 𝑎⨁𝑏 and
𝑐 ∈ 𝐺. This property is called closure property. A set following the closure property is called a group,
if –
• The binary operator is associative
• There is an identity element in 𝐺
 • There is an inverse element in 𝐺
A group wherein the binary operator is commutative is called an abelian. A group has certain theorem
attached to them –
• A group has a unique identity element
• A group has a unique inverse element
• In a group 𝐺, (𝑎−1 )−1 = 𝑎 ∀𝑎 ∈ 𝐺. Also, (𝑎𝑏)−1 = 𝑏 −1 𝑎−1.
The number of elements in the group is called its order. Minimum order value will be 1 which will be
the identity element. We can define some terms that help to get the bigger picture –
• Semi-group – This is a non-empty set that with a binary operation that follows the closure and
associative properties.
• Monoid – This is a semi-group that has an identity element.
• Group – It is a Monoid that has an inverse element.
• Abelian – It is a group which satisfies the commutative property.
• Cyclic Group – A group 𝐺 is said to be cyclic if there is an element 𝑥 ∈ 𝐺 such that all elements
in 𝐺 can be written as 𝑥 𝑛 where 𝑛 ∈ 𝑍. A cyclic group is always an abelian group.
• Sub-group – A subset 𝐺′ of a group 𝐺 is said to be a sub-group if 𝐺′ satisfies the properties of
a group. If 𝐺 is cyclic, then 𝐺′ will also be cyclic. If 𝐺 is abelian, then 𝐺′ will also be abelian.
• Partially Ordered Set (POSET) – This is a non-void set with a binary relation that is reflexive,
anti-symmetric and transitive.
More about POSETs, Hasse Diagram and Lattices here -
https://www.tutorialspoint.com/discrete_mathematics/discrete_mathematics_group_theory.htm

GRAPH THEORY
We already know a little bit about general graphs, vertices and edges. Here is some more points that
were not covered before.

Bi-partite Graph
A graph 𝐺 is bi-partite if it can be split into two disjoint sets in such a way that each edge has one end
in 1 set and the other end in the other set. If every node in one set is connected to every node in the
other set, then it is called a complete bi-partite graph.
Planar Graph
A graph which can be drawn and doesn’t have any edges crossing over each other is called a planar
graph.
Euler Graph
A connected graph 𝐺 is called an Euler graph, if there is a closed trail which includes every edge of the
graph 𝐺. An Euler path is a path that uses every edge of a graph exactly once. An Euler path starts and
ends at different vertices.
Graph Colouring
This is the process to assign a colour to each node in such a way that the neighbours have a different
colour. The minimum number of colours needed is called the chromatic number.
COMBINATORICS
This is the study of counting, combination and permutations.

BINOMIAL THEOREM
(𝑥 + 𝑦)𝑛 = 𝐶0𝑛 𝑥 𝑛 + 𝐶1𝑛 𝑥 𝑛−1 𝑦 + 𝐶2𝑛 𝑥 𝑛−2 𝑦 2 + ⋯ + 𝐶𝑛−1
𝑛
𝑥𝑦 𝑛−1 + 𝐶𝑛𝑛 𝑦 𝑛
This works when 𝑛 is a non-negative integer. This can be used to also write –

(1 + 𝑥)𝑛 = 𝐶0𝑛 + 𝐶1𝑛 𝑦 + 𝐶2𝑛 𝑦 2 + ⋯ + 𝐶𝑛−1

𝑛
𝑦 𝑛−1 + 𝐶𝑛𝑛 𝑦 𝑛

Question

Answer
25
𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 = 𝐶13 ∗ 213 ∗ (−3)12

NOTE

𝐶0𝑛 + 𝐶1𝑛 + 𝐶2𝑛 + ⋯ + 𝐶𝑛−1

𝑛
+ 𝐶𝑛𝑛 = 2𝑛

Suppose we have 𝑟 types of a product and we have to select 𝑛 products of any type. Then the number
of ways to do so is given by 𝑪𝒏+𝒓−𝟏
𝒏

Question

A man wants to buy 7 birds. The shop has pigeons, sparrows and finches. What is the total number of
ways for the man to buy the birds?

Answer

Here, 𝑛 = 7 ; 𝑟 = 3. Therefore,

𝑁 = 𝐶77+3−1 = 𝟑𝟔

DERANGEMENT

It is the situation where none of the objects are in their designated places. For 𝑛 objects, the
derangement is given by –
1 1 1 1 1
𝐷𝑛 = 𝑛! ∗ ( − + − + ⋯ + (−1)𝑛 )
2! 3! 4! 5! 𝑛!
Here, 𝑛 = 8. Thus, we get –

𝐷8 = 𝟏𝟒𝟖𝟑𝟑

GENERATING FUNCTIONS

For a series of numbers 𝐴 = [𝑎0 , 𝑎1 , 𝑎2 , … , 𝑎𝑛 ], the generating function can be written as –

𝑛

𝐺(𝑥) = ∑ 𝑎𝑖 𝑥 𝑖
𝑖=1

For example, let us take 𝐴 = [1,2,3, … ]. The generating function can be written as –

𝐺(𝑥) = 1 + 2𝑥 + 3𝑥 2 + 4𝑥 3 + ⋯

𝑥𝐺(𝑥) = 𝑥 + 2𝑥 2 + 3𝑥 3 + 4𝑥 4 + ⋯
Thus, we get –

𝐺(𝑥)(1 − 𝑥) = 1 + 𝑥 + 𝑥 2 + 𝑥 3 + ⋯

𝑥𝐺(𝑥)(1 − 𝑥) = 𝑥 + 𝑥 2 + 𝑥 3 + ⋯
Now, we get –

𝐺(𝑥)(1 − 𝑥)2 = 1
𝟏
𝑮(𝒙) =
(𝟏 − 𝒙)𝟐

TAYLOR SERIES

Taylor series is the polynomial or a function of an infinite sum of terms.

𝑓 ′ (𝑎) 𝑓 ′′ (𝑎) 𝑓 ′′′ (𝑎)
𝑓(𝑥) = 𝑓(𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎)2 + (𝑥 − 𝑎)3 + ⋯
1! 2! 3!
When 𝑎 = 0, the series is called Maclaurin series.
 𝑓 ′ (0) 𝑓 ′′ (0) 2 𝑓 ′′′ (0) 3
𝑓(𝑥) = 𝑓(0) + 𝑥+ 𝑥 + 𝑥 +⋯
1! 2! 3!

DIAGONALIZATION OF A MATRIX

Diagonalization is the process of writing a square matrix 𝐴 as follows –

𝐴 = 𝐶𝐷𝐶 −1
Where 𝐶 is the matrix where every column is an eigenvector of 𝐴 and 𝐷 is a diagonal matrix where the
diagonal elements are eigenvalues of 𝐴. The question now is, why even do this? Why even diagonalize
a square matrix? The answer to this is in the process of finding the power of diagonal matrix. Basically,
“the 𝑛𝑡ℎ power of a diagonal matrix is the same as taking the 𝑛𝑡ℎ power of each element.” For example,

1 0 10 10
010 ] = [1 0
[ ] = [110 10 ]
0 2 0 2 0 1024
Now, let’s find the value of 𝐴2 –

𝐴2 = 𝐴. 𝐴 = 𝐶𝐷𝐶 −1 . 𝐶𝐷𝐶 −1 = 𝐶𝐷𝐷𝐶 −1 = 𝐶𝐷 2 𝐶 −1

In short, thanks to diagonalization –

𝐴𝑛 = 𝐶𝐷 𝑛 𝐶 −1
It is now much easier to calculate higher powers of 𝐴. However, 𝐴 can only be diagonalized when 𝐶 is
invertible which in turn will only happen when it has 𝑛 linearly independent rows/columns. Therefore,
for a matrix 𝐴 to be diagonalizable, we need –

• To have 𝑛 linear independent eigenvectors

• To have 𝑛 distinct eigenvalues

Question

Find if the matrix is diagonalizable –

1 0 −1
𝐴 = [0 1 −1]
1 1 −2
Answer

We can see that for this matrix, 𝜆 = {0,1, −1}. We can now create the eigen vectors of the matrix as
follows –
1 −1 1
𝐶 = [1 1 1]
1 0 2
We can see that 𝐶 has 3 linearly independent rows and also |𝐶| = 2 ≠ 0. Therefore, 𝐶 is invertible and
hence 𝐴 is diagonalizable.
INTEGRATION TRICKS

Product of exponential and regular function

Let us assume we have a function as follows –

𝐼 = ∫ 𝑒 𝑔(𝑥) 𝑓(𝑥)𝑑𝑥

Let us assume 𝐷 = 𝑒 𝑔(𝑥) and 𝐹 = 𝑓(𝑥). Then, the integration can be solved as –

𝐼 = 𝐷𝐹 ′ − 𝐷 ′ 𝐹 ′′ + 𝐷 ′′ 𝐹 ′′′ − 𝐷 ′′′ 𝐹 ′′′′ + ⋯

This can be used when 𝐹 is an algebraic function. However, if 𝐹 is a trigonometric function, then there
is a problem because the function will never end as the differentials will start repeating. In such a case,
we can write –

𝐼 = 𝐷𝐹 ′ − 𝐷 ′ 𝐹 ′′ + ∫ 𝐷′′𝐹′′

Quadratic equation in the denominator

Suppose we have a function as follows –

𝑑𝑥
𝐼=∫
𝑎𝑥 2 + 𝑏𝑥 + 𝑐
To solve this, we first need to calculate the determinant of the quadratic equation. If 𝑫 > 𝟎, then –
1 𝑥∓𝛼
𝐼= log | |
√𝐷 𝑥∓𝛽
On the other hand, if 𝑫 < 𝟎, then –
2 𝐷𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑡𝑖𝑎𝑡𝑒 𝑡ℎ𝑒 𝑞𝑢𝑎𝑑𝑟𝑎𝑡𝑖𝑐 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛
𝐼= 𝑡𝑎𝑛−1 ( )
√−𝐷 √−𝐷

Question
𝑑𝑥
𝐼=∫
𝑥2 +𝑥+1
Answer

Here,

𝐷 = 𝑏 2 − 4𝑎𝑐 = −3 < 0
Hence,
2 2𝑥 + 1
𝐼= 𝑡𝑎𝑛−1 ( )
√3 √3
Question
𝑑𝑥
𝐼=∫
𝑥 2 + 3𝑥 + 2
Answer

Here,

𝐷 = 𝑏 2 − 4𝑎𝑐 = 1 > 0
Hence, we need to find the roots of the equation. We get –

𝛼 = 1 ;𝛽 = 2
Thus, we get –
1 𝑥+1
𝐼= log | |
√1 𝑥+2

Integration by parts

As per this trick, we have –

𝐼 = ∫ 𝑒 𝑥 [𝑓(𝑥) + 𝑓 ′ (𝑥)]𝑑𝑥 = 𝑒 𝑥 𝑓(𝑥)

TO WATCH AGAIN
• https://www.youtube.com/watch?v=HtuVQxzT73k&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZk
Az&index=20
• https://www.youtube.com/watch?v=HtuVQxzT73k&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZk
Az&index=21
• https://www.youtube.com/watch?v=HtuVQxzT73k&list=PLvTTv60o7qj_tdY9zH7YceES7jfXiZk
Az&index=22
•