Influencing policy (training slides from Fast Track Impact)
Orthogonality
1. Definition
Orthogonal Vectors
Two vectors u, v are said to be orthogonal provided their dot product is zero:
u·v = 0.
If both vectors are nonzero (not required in the definition), then the angle θ
between the two vectors is determined by
cos θ =
𝑢⋅𝑣
𝑢 𝑣
= 0,
which implies θ = 90◦.
In short, orthogonal vectors form a right angle.
2. Unitization/Normalization
Any nonzero vector u can be multiplied by c =
1
𝑢
to make a unit vector v = cu,
that is, a vector satisfying ||v|| = 1.
This process of changing the length of a vector to 1 by scalar multiplication is
called
Unitization/ Normalization.
3. Orthogonal and
Orthonormal vectors
A set {f1, f2, ..., fm} of non zero vectors in 𝑅𝑛 was called an orthogonal set
if fi·fj = 0 for all i ≠ j.
An orthogonal set is said to be orthonormal if each vector fm is a unit
vector.ie ||fm||=1
4. Independence and Orthogonality
Theorem 1 :An orthogonal set of nonzero
vectors is linearly independent.
Proof: Let c1, . . . , ck be constants such that nonzero orthogonal vectors u1, .
. . , uk satisfy the relation
c1u1 + · · · + ckuk = 0.
Take the dot product of this equation with vector uj to obtain the scalar relation
c1u1 · uj + · · · + ckuk · uj = 0.
Because all terms on the left are zero, except one, the relation reduces to the
simpler equation
Cj||uj||2 = 0.
This equation implies cj = 0. Therefore, c1 = · · · = ck = 0 and the vectors are
proved Linearly independent.
5. Projecion
Let V,W V and w=0. then a real number ‘c’ is called the
component of V along W if V-cw is orthogonal to W
ie <V-cw,w>=0
<v,w>-c<w,w>=0
c=
<v,w>
<w,w>
Thus if w is any
6. Orthogonality property
Let P and Q be two vectors in Rn , then the differences
of Q and its vector projection onto P is
orthogonal to P.
ie P is orthogonal to [Q-(
𝑄⋅𝑃
𝑃
𝑝
𝑃
]
7. Grahm Schmidt orthogonalization
process
Let P and Q be two non zero vectors then the above orthogonality property
can be used to construct a new vector.
R=[Q-(
𝑄⋅𝑃
𝑃
𝑝
𝑃
]
Which is orthogonal to P
This process of constructing orthogonal vector is known as Grahm Schmidt
orthogonalization process.