Zulip Chat Archive

Stream: Is there code for X?

Topic: How to prove a multiple integral is Integrable?


Junqi Liu (Jul 13 2024 at 04:10):

like

example : IntervalIntegrable
    (fun x   (y : ) (z : ) in (0)..1, 1 / (1 - x * y * z)) MeasureTheory.volume 0 1 := by sorry

where xx can't be separated from the multiple integral.
note: we assume this multiple integral must be integrable, maybe it is not 01010111xyzdxdydz\int_0^1 \int_0^1 \int_0^1 \frac{1}{1 - x * y * z}dx dy dz.

Yury G. Kudryashov (Jul 15 2024 at 00:56):

Do you have an informal proof?

Eric Wieser (Jul 15 2024 at 01:04):

This statement looks risky to me; if the inner integral is not integrable, then mathlib defines it as zero and so the outer one ends up integrable anyway!

Yury G. Kudryashov (Jul 15 2024 at 01:23):

The inner integral is well defined for all 0x<10 \le x < 1, because the function is continuous on [0,1]×[0,1][0, 1]\times [0, 1] in this case.

Eric Wieser (Jul 15 2024 at 01:27):

Right, in this case it's (presumably) true, but the statement is much weaker than the one that was intended.

Junqi Liu (Jul 15 2024 at 03:13):

:sob: My informal proof is that the triple integral is finite. But I don't know how to formal this proof.

Junqi Liu (Jul 15 2024 at 03:18):

Eric Wieser said:

This statement looks risky to me; if the inner integral is not integrable, then mathlib defines it as zero and so the outer one ends up integrable anyway!

yes! Actually I want to proof this MultiIntegral is positive. But if I want to use the lemma intervalIntegral.intervalIntegral_pos_of_pos_on. Then I need to show the Integral and MultiIntegral is Integrable.

Yury G. Kudryashov (Jul 15 2024 at 03:23):

Can you just compute the integral?

Junqi Liu (Jul 15 2024 at 03:43):

emmmmm. I can only compute a upper bound in informal space.

Yury G. Kudryashov (Jul 15 2024 at 04:22):

You can use docs#IntervalIntegrable.mono_fun'

Junqi Liu (Jul 15 2024 at 05:41):

ok! thanks~ let me have a try


Last updated: May 02 2025 at 03:31 UTC