Since f and g are both Riemann integrable, we have ∫If=∫If,∫Ig=∫Ig So for ∀ϵ>0, we have piecewise constant function f majorizes f, f minorizes f, g majorizes g, g minorizes g, such that ∫If−ϵ<∫If≤∫If≤∫If<∫If+ϵ,∫Ig−ϵ<∫Ig≤∫Ig≤∫Ig<∫Ig+ϵ ( a ) It’s easy to prove that f+g majorizes f+g and f+g minorizes f+g, so we have ∫I(f+g)≤∫I(f+g)≤∫I(f+g)≤∫I(f+g) Use Theorem 11.2.16 we have ∫I(f+g)=∫If+∫Ig>∫If+∫Ig−2ϵ∫I(f+g)=∫If+∫Ig<∫If+∫Ig+2ϵ Thus ∫I(f+g)−∫I(f+g)≤∫I(f+g)−∫I(f+g)<4ϵ So f+g is Riemann integrable, and ∫If+∫Ig−2ϵ<∫I(f+g)≤∫I(f+g)≤∫I(f+g)<∫If+∫Ig+2ϵ Which shows ∫I(f+g)=∫If+∫Ig. ( b ) First suppose c>0, then cf majorizes cf, cf minorizes cf, so we have ∫I(cf)≤∫I(cf)≤∫I(cf)≤∫I(cf) Use Theorem 11.2.16 we have ∫I(cf)=c∫If>c∫If−cϵ,∫I(cf)=c∫If<c∫If+cϵ Thus c∫If−cϵ<∫I(cf)≤∫I(cf)≤∫I(cf)≤∫I(cf)<c∫If+cϵ Which shows ∫I(cf)=c∫If. Now if c=0, then cf=0, which is a piecewise constant function on I, thus ∫I(cf)=p.c.∫I(cf)=p.c.∫I(0)=0=0⋅∫If Now suppose c=−1, then −f minorizes −f, −f majorizes −f, so we have ∫I(−f)≤∫I(−f)≤∫I(−f)≤∫I(−f) Use Theorem 11.2.16 we have ∫I(−f)=−∫If>−∫If−ϵ,∫I(−f)=−∫If<−∫If+ϵ Thus −∫If−ϵ<−∫If≤∫I(−f)≤∫I(−f)≤−∫If<−∫If+ϵ Which shows ∫I(−f)=−∫If. Finally for any c<0, we have c=(−1)(−c) and cf=(−1)(−cf),−c>0, so use the result for c=−1 and c>0, we have ∫I(cf)=∫I(−1)(−cf)=−∫I(−cf)=−(−c)∫If=c∫If ( c ) Use (a) and (b) we could have ∫I(f−g)=∫I(f+(−g))=∫If+∫I(−g)=∫If−∫Ig ( d ) The constant function h(x)=0 minorizes f, thus ∫If≥∫I(f)≥∫Ih=∫I0⋅h=0∫Ih=0 ( e ) The function f(x)−g(x)≥0,∀x∈I, thus use ( d ) and ( c ) ∫If−∫Ig=∫I(f−g)≥0⇒∫If≥∫Ig ( f ) In this case, we can use the piecewise constant function integral formula: ∫If=p.c.∫If=J∈P∑cJ∣J∣=cJ∈P∑∣J∣=c∣I∣ ( g ) We can define F and F which are piecewise and majorizes/minorizes F as: F(x)={f(x),0,x∈Ix∈/I,F(x)={f(x),0,x∈Ix∈/I Then use Theorem 11.2.16 we have ∫If−ϵ<∫If=∫JF≤∫JF≤∫JF≤∫JF=∫If<∫If+ϵ which shows ∫JF=∫If. ( h ) Use Theorem 11.2.16 (h) we have ∫If=∫Jf∣J+∫Kf∣K,∫If=∫Jf∣J+∫Kf∣K Thus ∫If−∫If=(∫Jf∣J−∫Jf∣J)+(∫Kf∣K−∫Kf∣K)<2ϵ Since we have f≥f≥f on I,J,K, this means ∫Jf∣J−∫Jf∣J≥0,∫Kf∣K−∫Kf∣K≥0 Thus we have ∫Jf∣J−∫Jf∣J<2ϵ,∫Kf∣K−∫Kf∣K<2ϵ Since f∣J majorizes f∣J and f∣J minorizes f∣J, we can say f∣J is Riemann integrable on J, by the same logic f∣K is Riemann integrable on K. We let F(x)={f(x),0,x∈Jx∈/J,G(x)={f(x),0,x∈Kx∈/K Then f(x)=F(x)+G(x), use (a) and (g) we have ∫If=∫IF+∫IG=∫Jf∣J+∫Kf∣K
Exercise 11.4.2
Assume ∃c∈[a,b],f(c)>0, since f is continuous, let ϵ=f(c)/2, then ∃δ>0, s.t. ∣f(x)−f(c)∣<ϵ=f(c)/2,∀x∈[a,b]∩(c−δ,c+δ) Since the length of [a,b]∩(c−δ,c+δ) is longer than δ/2, and we have f(x)>f(c)/2,∀x∈[a,b]∩(c−δ,c+δ) We can define g(x)={f(c)/2,f(x),x∈[a,b]∩(c−δ,c+δ)x∈[a,b],x∈/(c−δ,c+δ) Then we can say g minorizes f on [a,b], since f is Riemann integrable on [a,b], and {[a,b]∩(c−δ,c+δ),[a,b]\(c−δ,c+δ)} is a partition of [a,b], we can say the function f∣[a,b]\(c−δ,c+δ) is Riemann integrable by Theorem 11.4.1(h), and constant function is Riemann integrable by Theorem 11.4.1(f), thus g is integrable by repeated use of Theorem 11.4.1(g). Now use Theorem 11.4.1(h), (d),(f) we have 0=∫[a,b]f≥∫[a,b]g=∫[a,b]∩(c−δ,c+δ)(2f(c))+∫[a,b]\(c−δ,c+δ)f≥2f(c)∣∣∣∣2δ∣∣∣∣=4δf(c)>0 This is a contradiction.
Exercise 11.4.3
Repeated use of Theorem 11.4.1(h), we can get ∫If=J∈P∑∫Jf∣J=J∈P∑∫Jf
Exercise 11.4.4
It’s easy to see from Theorem 11.4.1 that if f is Riemann integrable, then so is –f. As min(f,g)=−max(−f,−g), we settled Theorem 11.4.3. For Theorem 11.4.5, we already have f+g+ Riemann integrable if f and g are integrable, notice that f+g−=f+(−g)+f−g+=(−f)+g+f−g−=(−f)+(−g)+ and both –f and –g are Riemann integrable, the conclusion follows.