Differential of the norm in $mathbbR^n$

Multi tool use
Clash Royale CLAN TAG#URR8PPP
up vote
1
down vote
favorite
We consider the normed vector space $(mathbbR^n,VertcdotVert)$.
Is the map $VertcdotVert$ differentiable even if it is not induced by a scalar product?
derivatives
add a comment |Â
up vote
1
down vote
favorite
We consider the normed vector space $(mathbbR^n,VertcdotVert)$.
Is the map $VertcdotVert$ differentiable even if it is not induced by a scalar product?
derivatives
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09
add a comment |Â
up vote
1
down vote
favorite
up vote
1
down vote
favorite
We consider the normed vector space $(mathbbR^n,VertcdotVert)$.
Is the map $VertcdotVert$ differentiable even if it is not induced by a scalar product?
derivatives
We consider the normed vector space $(mathbbR^n,VertcdotVert)$.
Is the map $VertcdotVert$ differentiable even if it is not induced by a scalar product?
derivatives
derivatives
asked Sep 10 at 11:52
net
62
62
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09
add a comment |Â
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09
add a comment |Â
3 Answers
3
active
oldest
votes
up vote
2
down vote
Sometimes yes, sometimes no. For example, consider the $p$-norms:
$$
|(x_1,dots,x_n)|_p = left(sum_k=1^n |x_k|^pright)^1/p .
$$
This is differentiable (except at the origin) if $p>1$, but not if $p=1$.
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
add a comment |Â
up vote
2
down vote
A norm on $mathbbR^n$ is never differentiable at the origin. Let $x_0in mathbbR^nsetminus left0right$. Then
$$lim_tto 0fract =lim_tto 0fract|x_0|$$
By the norm axiom $x_0neq 0 Rightarrow |x_0|neq 0$, and so the limit does not exist, as
$$lim_tto 0^+fract=1neq -1=lim_tto 0^-fract $$
On the other hand, the triangle inequality
$$left||x|-|y|right|leq |x-y| $$
tells us that any norm in $mathbbR^n$ is Lipschitz with Lipschitz constant $L=1$. Thus by Rademacher's theorem, it is differentiable almost everywhere in $mathbbR^n$.
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
add a comment |Â
up vote
0
down vote
A simple example of a norm on $Bbb R^2$ which is non-differentiable at points other than the origin is given by $$||(x,y)||=max(|x|,|y|).$$
If $h>0$ then $||(1+h,1)||=1+h$, while if $-2<h<0$ then $||(1+h,1)||=1$. Hence the function $hmapsto||(1+h,1)||$ is not differentiable at $h=0$, so $||.||$ is not differentiable at $(1,1)$.
add a comment |Â
3 Answers
3
active
oldest
votes
3 Answers
3
active
oldest
votes
active
oldest
votes
active
oldest
votes
up vote
2
down vote
Sometimes yes, sometimes no. For example, consider the $p$-norms:
$$
|(x_1,dots,x_n)|_p = left(sum_k=1^n |x_k|^pright)^1/p .
$$
This is differentiable (except at the origin) if $p>1$, but not if $p=1$.
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
add a comment |Â
up vote
2
down vote
Sometimes yes, sometimes no. For example, consider the $p$-norms:
$$
|(x_1,dots,x_n)|_p = left(sum_k=1^n |x_k|^pright)^1/p .
$$
This is differentiable (except at the origin) if $p>1$, but not if $p=1$.
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
add a comment |Â
up vote
2
down vote
up vote
2
down vote
Sometimes yes, sometimes no. For example, consider the $p$-norms:
$$
|(x_1,dots,x_n)|_p = left(sum_k=1^n |x_k|^pright)^1/p .
$$
This is differentiable (except at the origin) if $p>1$, but not if $p=1$.
Sometimes yes, sometimes no. For example, consider the $p$-norms:
$$
|(x_1,dots,x_n)|_p = left(sum_k=1^n |x_k|^pright)^1/p .
$$
This is differentiable (except at the origin) if $p>1$, but not if $p=1$.
edited Sep 10 at 11:58
answered Sep 10 at 11:57
GEdgar
59.2k265165
59.2k265165
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
add a comment |Â
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
@HenningMakholm: thanks, edited.
– GEdgar
Sep 10 at 11:59
1
1
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
It could also be remarked that even the 1-norm is differentiable almost everywhere -- namely at points where all the $x_i$s are nonzero.
– Henning Makholm
Sep 10 at 12:02
1
1
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
And, more generally, any convex function is differentiable almost everywhere?
– GEdgar
Sep 10 at 12:04
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
@GEdgar: yes math.stackexchange.com/q/727789/8157
– Giuseppe Negro
Sep 10 at 12:13
add a comment |Â
up vote
2
down vote
A norm on $mathbbR^n$ is never differentiable at the origin. Let $x_0in mathbbR^nsetminus left0right$. Then
$$lim_tto 0fract =lim_tto 0fract|x_0|$$
By the norm axiom $x_0neq 0 Rightarrow |x_0|neq 0$, and so the limit does not exist, as
$$lim_tto 0^+fract=1neq -1=lim_tto 0^-fract $$
On the other hand, the triangle inequality
$$left||x|-|y|right|leq |x-y| $$
tells us that any norm in $mathbbR^n$ is Lipschitz with Lipschitz constant $L=1$. Thus by Rademacher's theorem, it is differentiable almost everywhere in $mathbbR^n$.
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
add a comment |Â
up vote
2
down vote
A norm on $mathbbR^n$ is never differentiable at the origin. Let $x_0in mathbbR^nsetminus left0right$. Then
$$lim_tto 0fract =lim_tto 0fract|x_0|$$
By the norm axiom $x_0neq 0 Rightarrow |x_0|neq 0$, and so the limit does not exist, as
$$lim_tto 0^+fract=1neq -1=lim_tto 0^-fract $$
On the other hand, the triangle inequality
$$left||x|-|y|right|leq |x-y| $$
tells us that any norm in $mathbbR^n$ is Lipschitz with Lipschitz constant $L=1$. Thus by Rademacher's theorem, it is differentiable almost everywhere in $mathbbR^n$.
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
add a comment |Â
up vote
2
down vote
up vote
2
down vote
A norm on $mathbbR^n$ is never differentiable at the origin. Let $x_0in mathbbR^nsetminus left0right$. Then
$$lim_tto 0fract =lim_tto 0fract|x_0|$$
By the norm axiom $x_0neq 0 Rightarrow |x_0|neq 0$, and so the limit does not exist, as
$$lim_tto 0^+fract=1neq -1=lim_tto 0^-fract $$
On the other hand, the triangle inequality
$$left||x|-|y|right|leq |x-y| $$
tells us that any norm in $mathbbR^n$ is Lipschitz with Lipschitz constant $L=1$. Thus by Rademacher's theorem, it is differentiable almost everywhere in $mathbbR^n$.
A norm on $mathbbR^n$ is never differentiable at the origin. Let $x_0in mathbbR^nsetminus left0right$. Then
$$lim_tto 0fract =lim_tto 0fract|x_0|$$
By the norm axiom $x_0neq 0 Rightarrow |x_0|neq 0$, and so the limit does not exist, as
$$lim_tto 0^+fract=1neq -1=lim_tto 0^-fract $$
On the other hand, the triangle inequality
$$left||x|-|y|right|leq |x-y| $$
tells us that any norm in $mathbbR^n$ is Lipschitz with Lipschitz constant $L=1$. Thus by Rademacher's theorem, it is differentiable almost everywhere in $mathbbR^n$.
answered Sep 10 at 12:05


Lorenzo Quarisa
2,823316
2,823316
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
add a comment |Â
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
1
1
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
I like the reference to Rademacher's theorem but there is a subtlety. The theorem refers to the Euclidean norm on $mathbb R^n$, but this answer only shows that $lVertcdotrVert$ is 1-Lipschitz with respect to itself. To conclude that $lVert cdot rVert$ is Lipschitz with respect to the Euclidean norm ($ell^2$), we also need the inequality $$lVert xrVert le C lVert xrVert_ell^2.$$This inequality holds because all norms are equivalent on $mathbb R^n$.
– Giuseppe Negro
Sep 10 at 12:10
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
Absolutely. I noticed that I was working with different norms but since we are working in $mathbbR^n$ I did not worry about it :) but for a reader who is not familiar with this fact this is an important remark.
– Lorenzo Quarisa
Sep 10 at 12:17
add a comment |Â
up vote
0
down vote
A simple example of a norm on $Bbb R^2$ which is non-differentiable at points other than the origin is given by $$||(x,y)||=max(|x|,|y|).$$
If $h>0$ then $||(1+h,1)||=1+h$, while if $-2<h<0$ then $||(1+h,1)||=1$. Hence the function $hmapsto||(1+h,1)||$ is not differentiable at $h=0$, so $||.||$ is not differentiable at $(1,1)$.
add a comment |Â
up vote
0
down vote
A simple example of a norm on $Bbb R^2$ which is non-differentiable at points other than the origin is given by $$||(x,y)||=max(|x|,|y|).$$
If $h>0$ then $||(1+h,1)||=1+h$, while if $-2<h<0$ then $||(1+h,1)||=1$. Hence the function $hmapsto||(1+h,1)||$ is not differentiable at $h=0$, so $||.||$ is not differentiable at $(1,1)$.
add a comment |Â
up vote
0
down vote
up vote
0
down vote
A simple example of a norm on $Bbb R^2$ which is non-differentiable at points other than the origin is given by $$||(x,y)||=max(|x|,|y|).$$
If $h>0$ then $||(1+h,1)||=1+h$, while if $-2<h<0$ then $||(1+h,1)||=1$. Hence the function $hmapsto||(1+h,1)||$ is not differentiable at $h=0$, so $||.||$ is not differentiable at $(1,1)$.
A simple example of a norm on $Bbb R^2$ which is non-differentiable at points other than the origin is given by $$||(x,y)||=max(|x|,|y|).$$
If $h>0$ then $||(1+h,1)||=1+h$, while if $-2<h<0$ then $||(1+h,1)||=1$. Hence the function $hmapsto||(1+h,1)||$ is not differentiable at $h=0$, so $||.||$ is not differentiable at $(1,1)$.
answered Sep 10 at 14:34
David C. Ullrich
56k43787
56k43787
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%2f2911838%2fdifferential-of-the-norm-in-mathbbrn%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
Truth to be told, it's not differentiable (in $0$) when it's induced by a scalar product either.
– Saucy O'Path
Sep 10 at 11:57
A norm which is differentiable off the origin is called a "smooth norm".
– Theo Bendit
Sep 10 at 12:09