→ CNSaya :有反例 09/10 10:57
請問一下下面這件事會對嗎?
if f:[a,b]→R , g:[c,d]→R , f,g are absolutely continuous , g([c,d])<=[a,b]
(被包含)
then f(g(x)):[c,d]→R is absolutely continuous
我照定義做得時候:
for all e>0, there exists common d_e>0 s.t.
1.Σ│f(b_i)-f(a_i)│<e whenever Σ(b_i-a_i)<d_e and [a_i,b_i] nonoverlapping
2.Σ│g(d_i)-g(c_i)│<e whenever Σ(d_i-c_i)<d_e and [c_i,d_i] nonoverlapping
for 2. , take e = d_e , then 2. becomes:
Σ│g(d_i)-g(c_i)│<d_e when Σ(d_i-c_i)<d_(d_e) and [c_i,d_i] nonoverlapping
so take D_e=d_(d_e) , when Σ(d_i-c_i)<D_e and [c_i,d_i] nonoverlapping
we have Σ│f(g(d_i))-f(g(c_i))│ < e by 1. ????????????
目前無法辦到,因為雖然在Σ(d_i-c_i)<D_e下,我們有Σ│g(d_i)-g(c_i)│<e了
可是無法確定每個 [g(d_i),g(c_i)] 是 nonoverlapping,所以不能套入1.
有技巧可解決嗎?還是真的有反例?
謝謝
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 61.230.56.169