Prove that upper triangular matrices are closed under inversion
Clash Royale CLAN TAG#URR8PPP
up vote
2
down vote
favorite
Prove that group of upper triangle matrices $G$ is a subgroup of $GL_n(F)$
Contains identity: $I_nin G$ since $I_ij=0$ when $i>j$
Closed under multi: $C=AB, c_ij=sum_k=1^n=a_ikb_kj,i=1,2..,n,j=1,2..,n$ for all $c_ij,i>j$ either $ige k>j$ or $i>kge j$. One factor in each addend is $0$: either $a_ik=0$, or $b_kj=0$ thus $c_ij=0$ if $i>jrightarrow Cin G$
Closed under inversion: We know that a unique inverse exist since G is atleast a subset of $GL_n(F)$. Any tips on how do I proceed form here on? I thought about using my proof above to prove that $A^-1in G$ How do I know there is not a fringe case where $AA^-1=I, A^-1notin G$
abstract-algebra group-theory
add a comment |Â
up vote
2
down vote
favorite
Prove that group of upper triangle matrices $G$ is a subgroup of $GL_n(F)$
Contains identity: $I_nin G$ since $I_ij=0$ when $i>j$
Closed under multi: $C=AB, c_ij=sum_k=1^n=a_ikb_kj,i=1,2..,n,j=1,2..,n$ for all $c_ij,i>j$ either $ige k>j$ or $i>kge j$. One factor in each addend is $0$: either $a_ik=0$, or $b_kj=0$ thus $c_ij=0$ if $i>jrightarrow Cin G$
Closed under inversion: We know that a unique inverse exist since G is atleast a subset of $GL_n(F)$. Any tips on how do I proceed form here on? I thought about using my proof above to prove that $A^-1in G$ How do I know there is not a fringe case where $AA^-1=I, A^-1notin G$
abstract-algebra group-theory
1
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago
add a comment |Â
up vote
2
down vote
favorite
up vote
2
down vote
favorite
Prove that group of upper triangle matrices $G$ is a subgroup of $GL_n(F)$
Contains identity: $I_nin G$ since $I_ij=0$ when $i>j$
Closed under multi: $C=AB, c_ij=sum_k=1^n=a_ikb_kj,i=1,2..,n,j=1,2..,n$ for all $c_ij,i>j$ either $ige k>j$ or $i>kge j$. One factor in each addend is $0$: either $a_ik=0$, or $b_kj=0$ thus $c_ij=0$ if $i>jrightarrow Cin G$
Closed under inversion: We know that a unique inverse exist since G is atleast a subset of $GL_n(F)$. Any tips on how do I proceed form here on? I thought about using my proof above to prove that $A^-1in G$ How do I know there is not a fringe case where $AA^-1=I, A^-1notin G$
abstract-algebra group-theory
Prove that group of upper triangle matrices $G$ is a subgroup of $GL_n(F)$
Contains identity: $I_nin G$ since $I_ij=0$ when $i>j$
Closed under multi: $C=AB, c_ij=sum_k=1^n=a_ikb_kj,i=1,2..,n,j=1,2..,n$ for all $c_ij,i>j$ either $ige k>j$ or $i>kge j$. One factor in each addend is $0$: either $a_ik=0$, or $b_kj=0$ thus $c_ij=0$ if $i>jrightarrow Cin G$
Closed under inversion: We know that a unique inverse exist since G is atleast a subset of $GL_n(F)$. Any tips on how do I proceed form here on? I thought about using my proof above to prove that $A^-1in G$ How do I know there is not a fringe case where $AA^-1=I, A^-1notin G$
abstract-algebra group-theory
abstract-algebra group-theory
asked 3 hours ago
Wen2400
257
257
1
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago
add a comment |Â
1
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago
1
1
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago
add a comment |Â
4 Answers
4
active
oldest
votes
up vote
3
down vote
accepted
Suppose $A^-1$ is not upper triangular. Consider any one of the rows, say $i$, of $A^-1$ such that it has non - zero entries before it's diagonal entry $a_ii$. Let the first of these be in the $j^th$ column.
Finally, a hint - consider the element of the product of $A^-1A$ in the position $(i,j)$
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
add a comment |Â
up vote
2
down vote
If $A$ is an invertible matrix, then its inverse $A^-1$ can be expressed as a polynomial in $A$.
Therefore, if $A$ is upper triangular, so is $A^-1$, since the product of upper triangular matrices is upper triangular, as you have already proved.
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
add a comment |Â
up vote
1
down vote
Conceptually, the algebra of upper triangular matrices $UT_n(F)$ is the set of transformations which fixes a complete flag of subspaces of $F^n$.
That is, for all upper triangular matrices, there is a chain of subspaces of $F^n$ written as $0subseteq V_1subseteq V_2subseteqldotssubseteq V_n$, where $dim_F(V_i)=i$, such that for every upper triangular matrix $A$, and each index $i$, $A(V_i)subseteq V_i$. Conversely given any such flag, you can choose a basis such that the flag-preserving transformations are upper triangular. You should have no problem describing such a flag of subspaces of $F^n$ such that the transformations are the upper triangular matrices.
If $A$ is invertible, then it preserves dimension, so $A(V_i)=V_i$ for all $i$. The inverse map, if it exists, would map $V_i$ right back into $V_i$, so the inverse is also an upper triangular matrix.
add a comment |Â
up vote
0
down vote
Thanks for all the help, I will now attempt to solve this using AnotherJohnDoe's hint. Any pointers or feedback to make this proof more elegant are appreciated.
Let $BA=I,Ain G$ then I choose to look at entries $I_i1,i=2,3,..$ then $I_i1=0=sum_k=1^nb_ika_k1$ since $a_k1=0,k>1$ the expression reduces to $I_i1=0=b_i1a_11$. Determinant of an upper triangle matrix is the product of the diagonal entries, in our case none of the diagonal entries can be zero. Thus $b_i1=0$.
Moving on to $I_i2,i=3,4...$ then $I_i2=0=sum_k=1^nb_ika_k2$ since $a_k2=0,k>2$ the expression reduces to $I_i2=0=b_i1a_12+b_i2a_22$ since $b_i1=0,a_22neq0$ Only solution is $b_i2=0$
Now using induction to complete the proof. (Which I might give it a go at a later date)
add a comment |Â
4 Answers
4
active
oldest
votes
4 Answers
4
active
oldest
votes
active
oldest
votes
active
oldest
votes
up vote
3
down vote
accepted
Suppose $A^-1$ is not upper triangular. Consider any one of the rows, say $i$, of $A^-1$ such that it has non - zero entries before it's diagonal entry $a_ii$. Let the first of these be in the $j^th$ column.
Finally, a hint - consider the element of the product of $A^-1A$ in the position $(i,j)$
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
add a comment |Â
up vote
3
down vote
accepted
Suppose $A^-1$ is not upper triangular. Consider any one of the rows, say $i$, of $A^-1$ such that it has non - zero entries before it's diagonal entry $a_ii$. Let the first of these be in the $j^th$ column.
Finally, a hint - consider the element of the product of $A^-1A$ in the position $(i,j)$
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
add a comment |Â
up vote
3
down vote
accepted
up vote
3
down vote
accepted
Suppose $A^-1$ is not upper triangular. Consider any one of the rows, say $i$, of $A^-1$ such that it has non - zero entries before it's diagonal entry $a_ii$. Let the first of these be in the $j^th$ column.
Finally, a hint - consider the element of the product of $A^-1A$ in the position $(i,j)$
Suppose $A^-1$ is not upper triangular. Consider any one of the rows, say $i$, of $A^-1$ such that it has non - zero entries before it's diagonal entry $a_ii$. Let the first of these be in the $j^th$ column.
Finally, a hint - consider the element of the product of $A^-1A$ in the position $(i,j)$
answered 2 hours ago
AnotherJohnDoe
824212
824212
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
add a comment |Â
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
It seems like I can construct a combination of numbers where non zero terms cancel eachother out. Therefore i was reluctant using similair proof as I did above
â Wen2400
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
But, in particular, what happens to the element in position $(i,j)?$
â AnotherJohnDoe
2 hours ago
add a comment |Â
up vote
2
down vote
If $A$ is an invertible matrix, then its inverse $A^-1$ can be expressed as a polynomial in $A$.
Therefore, if $A$ is upper triangular, so is $A^-1$, since the product of upper triangular matrices is upper triangular, as you have already proved.
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
add a comment |Â
up vote
2
down vote
If $A$ is an invertible matrix, then its inverse $A^-1$ can be expressed as a polynomial in $A$.
Therefore, if $A$ is upper triangular, so is $A^-1$, since the product of upper triangular matrices is upper triangular, as you have already proved.
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
add a comment |Â
up vote
2
down vote
up vote
2
down vote
If $A$ is an invertible matrix, then its inverse $A^-1$ can be expressed as a polynomial in $A$.
Therefore, if $A$ is upper triangular, so is $A^-1$, since the product of upper triangular matrices is upper triangular, as you have already proved.
If $A$ is an invertible matrix, then its inverse $A^-1$ can be expressed as a polynomial in $A$.
Therefore, if $A$ is upper triangular, so is $A^-1$, since the product of upper triangular matrices is upper triangular, as you have already proved.
answered 49 mins ago
lhf
158k9161375
158k9161375
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
add a comment |Â
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
When I did some testing on $3times 3$ matrix, I found that $I_32=0=a_31^-1a_12+a_32^-1a_22+a_33^-1a_32$ reduces to $0=a_31^-1a_12+a_32^-1a_22$ Which got me worried that somewhere out there there might exist a fringe case where $a_31^-1,a_32^-1neq 0$
â Wen2400
32 mins ago
+1 Even better!
â rschwieb
7 mins ago
+1 Even better!
â rschwieb
7 mins ago
add a comment |Â
up vote
1
down vote
Conceptually, the algebra of upper triangular matrices $UT_n(F)$ is the set of transformations which fixes a complete flag of subspaces of $F^n$.
That is, for all upper triangular matrices, there is a chain of subspaces of $F^n$ written as $0subseteq V_1subseteq V_2subseteqldotssubseteq V_n$, where $dim_F(V_i)=i$, such that for every upper triangular matrix $A$, and each index $i$, $A(V_i)subseteq V_i$. Conversely given any such flag, you can choose a basis such that the flag-preserving transformations are upper triangular. You should have no problem describing such a flag of subspaces of $F^n$ such that the transformations are the upper triangular matrices.
If $A$ is invertible, then it preserves dimension, so $A(V_i)=V_i$ for all $i$. The inverse map, if it exists, would map $V_i$ right back into $V_i$, so the inverse is also an upper triangular matrix.
add a comment |Â
up vote
1
down vote
Conceptually, the algebra of upper triangular matrices $UT_n(F)$ is the set of transformations which fixes a complete flag of subspaces of $F^n$.
That is, for all upper triangular matrices, there is a chain of subspaces of $F^n$ written as $0subseteq V_1subseteq V_2subseteqldotssubseteq V_n$, where $dim_F(V_i)=i$, such that for every upper triangular matrix $A$, and each index $i$, $A(V_i)subseteq V_i$. Conversely given any such flag, you can choose a basis such that the flag-preserving transformations are upper triangular. You should have no problem describing such a flag of subspaces of $F^n$ such that the transformations are the upper triangular matrices.
If $A$ is invertible, then it preserves dimension, so $A(V_i)=V_i$ for all $i$. The inverse map, if it exists, would map $V_i$ right back into $V_i$, so the inverse is also an upper triangular matrix.
add a comment |Â
up vote
1
down vote
up vote
1
down vote
Conceptually, the algebra of upper triangular matrices $UT_n(F)$ is the set of transformations which fixes a complete flag of subspaces of $F^n$.
That is, for all upper triangular matrices, there is a chain of subspaces of $F^n$ written as $0subseteq V_1subseteq V_2subseteqldotssubseteq V_n$, where $dim_F(V_i)=i$, such that for every upper triangular matrix $A$, and each index $i$, $A(V_i)subseteq V_i$. Conversely given any such flag, you can choose a basis such that the flag-preserving transformations are upper triangular. You should have no problem describing such a flag of subspaces of $F^n$ such that the transformations are the upper triangular matrices.
If $A$ is invertible, then it preserves dimension, so $A(V_i)=V_i$ for all $i$. The inverse map, if it exists, would map $V_i$ right back into $V_i$, so the inverse is also an upper triangular matrix.
Conceptually, the algebra of upper triangular matrices $UT_n(F)$ is the set of transformations which fixes a complete flag of subspaces of $F^n$.
That is, for all upper triangular matrices, there is a chain of subspaces of $F^n$ written as $0subseteq V_1subseteq V_2subseteqldotssubseteq V_n$, where $dim_F(V_i)=i$, such that for every upper triangular matrix $A$, and each index $i$, $A(V_i)subseteq V_i$. Conversely given any such flag, you can choose a basis such that the flag-preserving transformations are upper triangular. You should have no problem describing such a flag of subspaces of $F^n$ such that the transformations are the upper triangular matrices.
If $A$ is invertible, then it preserves dimension, so $A(V_i)=V_i$ for all $i$. The inverse map, if it exists, would map $V_i$ right back into $V_i$, so the inverse is also an upper triangular matrix.
answered 55 mins ago
rschwieb
101k1198234
101k1198234
add a comment |Â
add a comment |Â
up vote
0
down vote
Thanks for all the help, I will now attempt to solve this using AnotherJohnDoe's hint. Any pointers or feedback to make this proof more elegant are appreciated.
Let $BA=I,Ain G$ then I choose to look at entries $I_i1,i=2,3,..$ then $I_i1=0=sum_k=1^nb_ika_k1$ since $a_k1=0,k>1$ the expression reduces to $I_i1=0=b_i1a_11$. Determinant of an upper triangle matrix is the product of the diagonal entries, in our case none of the diagonal entries can be zero. Thus $b_i1=0$.
Moving on to $I_i2,i=3,4...$ then $I_i2=0=sum_k=1^nb_ika_k2$ since $a_k2=0,k>2$ the expression reduces to $I_i2=0=b_i1a_12+b_i2a_22$ since $b_i1=0,a_22neq0$ Only solution is $b_i2=0$
Now using induction to complete the proof. (Which I might give it a go at a later date)
add a comment |Â
up vote
0
down vote
Thanks for all the help, I will now attempt to solve this using AnotherJohnDoe's hint. Any pointers or feedback to make this proof more elegant are appreciated.
Let $BA=I,Ain G$ then I choose to look at entries $I_i1,i=2,3,..$ then $I_i1=0=sum_k=1^nb_ika_k1$ since $a_k1=0,k>1$ the expression reduces to $I_i1=0=b_i1a_11$. Determinant of an upper triangle matrix is the product of the diagonal entries, in our case none of the diagonal entries can be zero. Thus $b_i1=0$.
Moving on to $I_i2,i=3,4...$ then $I_i2=0=sum_k=1^nb_ika_k2$ since $a_k2=0,k>2$ the expression reduces to $I_i2=0=b_i1a_12+b_i2a_22$ since $b_i1=0,a_22neq0$ Only solution is $b_i2=0$
Now using induction to complete the proof. (Which I might give it a go at a later date)
add a comment |Â
up vote
0
down vote
up vote
0
down vote
Thanks for all the help, I will now attempt to solve this using AnotherJohnDoe's hint. Any pointers or feedback to make this proof more elegant are appreciated.
Let $BA=I,Ain G$ then I choose to look at entries $I_i1,i=2,3,..$ then $I_i1=0=sum_k=1^nb_ika_k1$ since $a_k1=0,k>1$ the expression reduces to $I_i1=0=b_i1a_11$. Determinant of an upper triangle matrix is the product of the diagonal entries, in our case none of the diagonal entries can be zero. Thus $b_i1=0$.
Moving on to $I_i2,i=3,4...$ then $I_i2=0=sum_k=1^nb_ika_k2$ since $a_k2=0,k>2$ the expression reduces to $I_i2=0=b_i1a_12+b_i2a_22$ since $b_i1=0,a_22neq0$ Only solution is $b_i2=0$
Now using induction to complete the proof. (Which I might give it a go at a later date)
Thanks for all the help, I will now attempt to solve this using AnotherJohnDoe's hint. Any pointers or feedback to make this proof more elegant are appreciated.
Let $BA=I,Ain G$ then I choose to look at entries $I_i1,i=2,3,..$ then $I_i1=0=sum_k=1^nb_ika_k1$ since $a_k1=0,k>1$ the expression reduces to $I_i1=0=b_i1a_11$. Determinant of an upper triangle matrix is the product of the diagonal entries, in our case none of the diagonal entries can be zero. Thus $b_i1=0$.
Moving on to $I_i2,i=3,4...$ then $I_i2=0=sum_k=1^nb_ika_k2$ since $a_k2=0,k>2$ the expression reduces to $I_i2=0=b_i1a_12+b_i2a_22$ since $b_i1=0,a_22neq0$ Only solution is $b_i2=0$
Now using induction to complete the proof. (Which I might give it a go at a later date)
answered 1 hour ago
Wen2400
257
257
add a comment |Â
add a comment |Â
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2938222%2fprove-that-upper-triangular-matrices-are-closed-under-inversion%23new-answer', 'question_page');
);
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
1
Try with the expression of $A^-1$ in term of its comatrix.
â AlexL
2 hours ago
Is the zero matrix also an upper triangular matrix?
â dan_fulea
2 hours ago
Yes, I believe so.
â AnotherJohnDoe
2 hours ago
Intuitively, in performing Gaussian elimination to convert the augmented matrix $beginbmatrix A & | & I endbmatrix$ to $beginbmatrix I & | & A^-1 endbmatrix$, you can restrict to operations which keep both halves of the augmented matrix upper triangular.
â Daniel Schepler
1 hour ago