0
$\begingroup$

Let $F_1, F_2$ - two continuous CDF.

if $F_1 = F_2\quad F_2$ almost surely (i.e. probability of $x$ where $F_1(x)\neq F_2(x)$ is zero with respect to probability with CDF $F_2$).

Then $F_1 = F_2$ (everywhere).

$\endgroup$
2
  • 1
    $\begingroup$ Hint: consider the contrapositive. If the $F_i$ are continuous and there exists $x$ for which $F_1(x)\ne F_2(x),$ can you show ${\Pr}_{F_2}(F_1 \ne F_2) \gt 0$? $\endgroup$
    – whuber
    Jan 22 at 15:50
  • 1
    $\begingroup$ Thank you, reformulation of statement helps. $\endgroup$ Jan 22 at 16:08
1
$\begingroup$

By contrapositive, if exists $x$ such that $F_1(x) \neq F_2(x)$

if $F_2(x) < F_1(x)$, then choose $y>x$ such that $F_2(y) > F_2(x)$ and $F_2(y) < F_1(x)$. Then $F_2(x) \neq F_1(x)$ on $[x,y]$ (by the monotony of F_1) and ${\Pr}_{F_2}([x,y]) \ge F_2(y) - F_2(x) > 0$.

if $F_2(x) > F_1(x)$, then choose $y<x$ such that $F_2(y) < F_2(x)$ and $F_2(y) > F_1(x)$. Then $F_2(x) \neq F_1(x)$ on $[y,x]$ (by the monotony of F_1) and ${\Pr}_{F_2}([y,x]) \ge F_2(x) - F_2(y)$ > 0.

Note I use only continuity of $F_2$ and statement true if $F_1$ is arbitrary.

$\endgroup$
1
  • $\begingroup$ +1: the final remark is insightful. $\endgroup$
    – whuber
    Jan 22 at 16:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.