Question about orthogonal projections.

In summary, oblique (non-orthogonal) projections map a vector back to itself and have certain properties such as a direct sum of subspaces.
  • #1
evilpostingmong
339
0
Aren't all projections orthogonal projections? What I mean is that let's say there
is a vector in 3d space and it gets projected to 2d space. So [1 2 3]--->[1 2 0]
Within the null space is [0 0 3], which is perpendicular to every vector in the x-y plane,
not to mention the inner product of [0 0 3] and (column)[1 2 0] is 0, which shows
that [0 0 3] is perpendicular to [1 2 0]. What also gets me is that I have seen
a picture of two vectors and one vector projected (orthogonally) onto the
other vector, but both were on the x-y plane, yet orthogonality is shown
by taking the inner product between two vectors and getting 0. I can't see
how 0 can be obtained when both vectors are on the same plane (note that they
were in the positive x positive y quadrant).
 
Physics news on Phys.org
  • #3
Let W be an underlying vector space. Suppose the subspaces U and V are the range and null space of P respectively. Then we have these basic properties:

1. P is the identity operator I on U: \forall x \in U: Px = x.
2. We have a direct sum W = U ⊕ V. This means that every vector x may be decomposed uniquely in the manner x = u + v, where u is in U and v is in V. The decomposition is given by u = Px,\ v = x - Px.

So by 1., oblique projections map the vector back to itself Pv=v right?
 
  • #4
evilpostingmong said:
Let W be an underlying vector space. Suppose the subspaces U and V are the range and null space of P respectively. Then we have these basic properties:

1. P is the identity operator I on U: \forall x \in U: Px = x.
2. We have a direct sum W = U ⊕ V. This means that every vector x may be decomposed uniquely in the manner x = u + v, where u is in U and v is in V. The decomposition is given by u = Px,\ v = x - Px.

So by 1., oblique projections map the vector back to itself Pv=v right?

(just got up … :zzz:)

Sorry, not following you :redface:

v is in V, and 1. only applies to U. :confused:
 
  • #5
Forget what I said. I was mistaken. But thanks for the link, tiny-tim!
 

1. What is an orthogonal projection?

An orthogonal projection is a type of mathematical transformation that projects a vector or point onto a subspace in a way that preserves the angle between the vector and the subspace. In simpler terms, it is a way to "project" an object onto a flat surface while keeping the object's shape intact.

2. How is an orthogonal projection different from other types of projections?

An orthogonal projection is unique in that it preserves the length and direction of the vector being projected. This means that the projected vector is always perpendicular to the subspace it is being projected onto. Other types of projections, such as parallel and oblique projections, do not have this property.

3. What are some real-world applications of orthogonal projections?

Orthogonal projections have many practical applications in fields such as engineering, computer graphics, and physics. They are used for creating 2D and 3D drawings, transforming data in statistics and data analysis, and in the construction of buildings and bridges.

4. How is an orthogonal projection calculated?

The calculation for an orthogonal projection involves finding the dot product between the vector being projected and a unit vector (a vector with a length of 1) in the direction of the subspace. This dot product is then multiplied by the unit vector to get the projected vector. The formula for an orthogonal projection is: projv(u) = (u * v / ||v||2) * v.

5. Can orthogonal projections be used in higher dimensions?

Yes, orthogonal projections can be used in any number of dimensions. The formula for an orthogonal projection remains the same, but the calculations become more complex as the number of dimensions increases. Orthogonal projections are commonly used in 3D space, but they can also be applied to higher-dimensional spaces in fields such as quantum mechanics and computer graphics.

Similar threads

Replies
3
Views
1K
  • Linear and Abstract Algebra
Replies
9
Views
566
Replies
2
Views
779
  • Linear and Abstract Algebra
Replies
14
Views
2K
Replies
26
Views
2K
  • Linear and Abstract Algebra
Replies
19
Views
1K
  • Linear and Abstract Algebra
Replies
9
Views
869
  • Linear and Abstract Algebra
Replies
4
Views
870
  • Linear and Abstract Algebra
Replies
3
Views
935
  • Linear and Abstract Algebra
Replies
10
Views
2K
Back
Top