High School Orthogonal Projections: Same Thing or Not?

Click For Summary
The discussion centers on whether orthogonal projections and general projections are the same concept in linear algebra. Participants clarify that a projection is defined by the property A^2 = A, and emphasize that A typically represents a matrix whose columns form a basis for the subspace, not the projection matrix itself. There is confusion regarding the simplification of matrix equations involving non-square matrices, highlighting that such simplifications are not valid. The conversation concludes with the understanding that if matrices were invertible, the projection would merely return the original vector in R^n. Overall, the distinction between orthogonal projections and general projections is affirmed.
Isaac0427
Insights Author
Gold Member
Messages
718
Reaction score
163
TL;DR
Why is the orthogonal projection formula written as ##P_A=A(A^TA)^{-1}A^T## as opposed to ##P_A=(A^T)^{-1}A^T##?
Physics news on Phys.org
##\left(A^\tau\right)^{-1}\left(A\right)^\tau=1## which doesn't mean a lot. ##A\left(A^\tau A\right)^{-1}A^\tau=1## which is also meaningless. A projection is generally a transformation for which ##A^2=A## holds.
 
You can't simplify ##(A^TA)^{-1}=A^{-1}(A^T)^{-1}## since ##A## is not a square matrix.

@fresh_42 I think ##A## is a matrix whose columns form a basis for the subspace of the projection, not the projection matrix itself.
 
Infrared said:
You can't simplify (ATA)−1=A−1(AT)−1 since A is not a square matrix.
Ah, that's what I was missing. Ok, thank you.
 
Infrared said:
@fresh_42 I think A is a matrix whose columns form a basis for the subspace of the projection, not the projection matrix itself.
And yes. I see though that it would all be meaningless if they were invertible matrices, as then the span of A would be all of ##\mathbb{R}^n##, and a vector in ##\mathbb{R}^n##'s projection onto ##\mathbb{R}^n## is just itself.
 
Thread 'How to define a vector field?'
Hello! In one book I saw that function ##V## of 3 variables ##V_x, V_y, V_z## (vector field in 3D) can be decomposed in a Taylor series without higher-order terms (partial derivative of second power and higher) at point ##(0,0,0)## such way: I think so: higher-order terms can be neglected because partial derivative of second power and higher are equal to 0. Is this true? And how to define vector field correctly for this case? (In the book I found nothing and my attempt was wrong...

Similar threads

  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 16 ·
Replies
16
Views
3K
Replies
16
Views
3K
  • · Replies 2 ·
Replies
2
Views
1K
Replies
10
Views
10K
  • · Replies 7 ·
Replies
7
Views
3K
Replies
13
Views
2K
Replies
20
Views
5K
  • · Replies 41 ·
2
Replies
41
Views
6K
Replies
25
Views
2K