When is (S o T) invertible? (Linear Maps)

  • Thread starter JoshMaths
  • Start date
In summary, for linear maps S and T: U -> V and V -> W, respectively, the composition SoT is invertible if and only if kerT = {0}, ImS = W, ImT ∩ kerS = {0}, and ImT (+) kerS = V. This can be understood by considering the properties of surjectivity and injectivity of the individual maps and their composition, as well as the rank-nullity theorem. It is important to note that the kernel of one map may not necessarily be zero, but it is the intersection of the images of both maps that ultimately determines invertibility.
  • #1
JoshMaths
26
0

Homework Statement


Show SoT is invertible iff
kerT = {0}
ImS = W
ImT n kerS = {0} and ImtT (+) kerS = V

T: U -> V
S: V -> W

The Attempt at a Solution



We know if kerSoT = {0} and kerT C kerSoT then kerT = {0}

ImS = W is implying surjectivity?
SoT = S(Tu) = Sv = W = ImS for all v (belongs to) ImT

I know the last part is stating the rank-nullity theorem but in a way I have not seen. If the kernel of S is zero then doesn't ImT = V or is kerS not necessarily zero?

I am also have difficulty with the whole premise of using set notations to explain properties of linear maps so if you could provide some guidance on this that would be great.

Josh
 
Physics news on Phys.org
  • #2
JoshMaths said:

Homework Statement


Show SoT is invertible iff
kerT = {0}
ImS = W
ImT ∩ kerS = {0} and ImT (+) kerS = V

T: U -> V
S: V -> W

The Attempt at a Solution



We know if kerSoT = {0} and kerT ⊂ kerSoT then kerT = {0}

ImS = W is implying surjectivity?
Depends on which mapping you're talking about. It tells you S is surjective, but S○T may not necessarily be. For example, if T mapped everything to 0, the image of S○T would be {0}.

SoT = S(Tu) = Sv = W = ImS for all v (belongs to) ImT
Your notation is a mess. You're equating a mapping (S○T) to vectors (S(Tu) and Sv) to a vector space (W).

I know the last part is stating the rank-nullity theorem but in a way I have not seen. If the kernel of S is zero then doesn't ImT = V or is kerS not necessarily zero?
Suppose S: ℝ3→ℝ2 maps (x,y,z) to (x,y), and T: ℝ2→ℝ3 maps (x,y) to (x,y,0). The kernel of S is the z-axis, but S○T is invertible.

I am also have difficulty with the whole premise of using set notations to explain properties of linear maps so if you could provide some guidance on this that would be great.

Josh
 

FAQ: When is (S o T) invertible? (Linear Maps)

1. When is (S o T) invertible?

The composition of two linear maps, (S o T), is invertible when both individual maps, S and T, are invertible and their composition results in a linear map that is one-to-one and onto.

2. Can (S o T) be invertible if S and T are not invertible?

No, (S o T) can only be invertible if both S and T are invertible. If either map is not invertible, their composition will also not be invertible.

3. How can I determine if S and T are invertible?

S and T are invertible if they have a nonzero determinant. This can be determined by calculating the determinant of their respective matrices, or by checking if they have a trivial/null kernel.

4. Are there any other conditions for (S o T) to be invertible?

In addition to S and T being invertible, another condition for (S o T) to be invertible is that the dimensions of the vector spaces of S and T must match. In other words, the output of T must be the same dimension as the input of S.

5. What is the significance of (S o T) being invertible?

When (S o T) is invertible, it means that the two maps S and T can be "undone" or reversed, and the original input can be obtained from the output. This is useful in many applications, such as solving systems of linear equations and transforming between different coordinate systems.

Similar threads

Back
Top