Equality holds in Minkowski's Inequality when


by benorin
Tags: equality, holds, inequality, minkowski
benorin
benorin is offline
#1
Jan11-06, 07:19 AM
HW Helper
P: 1,007
I would like to determine necessary and sufficient conditions for equality to hold in Minkowski's Inequality in [itex]L^{p}(X,\mu)[/itex].

For [itex]1\leq p\leq \infty,[/itex] we have [itex]\forall f,g\in L^{p}(X,\mu)[/itex]

[tex]\left\{\int_X \left| f+g\right| ^{p} d\mu\right\} ^{\frac{1}{p}} \leq \left\{\int_X \left| f\right| ^{p} d\mu\right\} ^{\frac{1}{p}} + \left\{\int_X \left| g\right| ^{p} d\mu\right\} ^{\frac{1}{p}}[/tex]

here I wish to allow f and g to be complex. Any help would be nice.
Phys.Org News Partner Science news on Phys.org
Better thermal-imaging lens from waste sulfur
Hackathon team's GoogolPlex gives Siri extra powers
Bright points in Sun's atmosphere mark patterns deep in its interior
benorin
benorin is offline
#2
Jan12-06, 06:20 AM
HW Helper
P: 1,007
So I think I have all the pieces, but I lack your brain power.

Minkowski's Inequality is proved (in Rudin's text) via Holder's Inequality, namely

For [itex]\frac{1}{p}+\frac{1}{q}=1[/itex], [itex]1< p< \infty,1<q< \infty,[/itex], and let f and g be measurable functions on X with range in [itex]\left[ 0, \infty\right] [/itex]. Then

[tex]\left\{\int_X fg d\mu\right\}\leq \left\{\int_X f^{p} d\mu\right\} ^{\frac{1}{p}}\left\{\int_X g^{q} d\mu\right\} ^{\frac{1}{q}}[/tex]

To prove Minkowski's put [itex](f+g)^{p}=f(f+g)^{p-1}+g(f+g)^{p-1}[/itex] and apply Holder's inequality to each of the terms on the right, that is

[tex]\left\{\int_X f(f+g)^{p-1} d\mu\right\}\leq \left\{\int_X f^{p} d\mu\right\} ^{\frac{1}{p}}\left\{\int_X (f+g)^{(p-1)q} d\mu\right\} ^{\frac{1}{q}}[/tex]

and

[tex]\left\{\int_X g(f+g)^{p-1} d\mu\right\}\leq \left\{\int_X g^{p} d\mu\right\} ^{\frac{1}{p}}\left\{\int_X (f+g)^{(p-1)q} d\mu\right\} ^{\frac{1}{q}}[/tex]

note that (p-1)q=p and that 1/q=1-1/p then add the above inequalities to get Minkowski's inequality in the following form:

[tex]\left\{\int_X \left( f+g\right) ^{p} d\mu\right\} ^{\frac{1}{p}} \leq \left\{\int_X f^{p} d\mu\right\} ^{\frac{1}{p}} + \left\{\int_X g^{p} d\mu\right\} ^{\frac{1}{p}}[/tex]

for non-negative f and g. Equality holds in Holder's inequality if, and only if, there are constants a and b such that [itex]af^p=bg^q[/itex].

To arrive at the form of Minkowski's inequality I have posted above, note that

[tex]\left| f + g \right| \leq \left| f \right| + \left| g \right| \Rightarrow \left| f + g \right| ^{p} \leq \left( \left| f \right| + \left| g \right| \right) ^{p}[/tex]

by the triangle inequality. For complex f and g, their magnitudes |f| and |g| are non-negative and measurable if f and g are (so it is ok to generalize as such.) Based on the conditions for equality to hold in Holder's Inequality, what are the conditions for equality to hold in Minkowski's Ineqaulity for complex f and g?
fourier jr
fourier jr is offline
#3
Jan12-06, 09:17 AM
P: 943
when one function is a constant multiple of the other, ie f = ag or somethng like that


Register to reply

Related Discussions
Minkowski's inequality! Calculus 2
Minkowski's Inequality Calculus & Beyond Homework 6
minkowski's 4 dimensional world Special & General Relativity 30
triangle inequality, parallelogram equality Precalculus Mathematics Homework 4
Minkowski's geometric Special & General Relativity 1