Zulip Chat Archive

Stream: Is there code for X?

Topic: e as limit of (1+1/n)^n


Bhavik Mehta (Jun 18 2021 at 00:26):

Does mathlib have a statement like this?

import analysis.specific_limits
import analysis.calculus.deriv
import analysis.special_functions.exp_log
import analysis.special_functions.pow

open_locale big_operators filter topological_space

example (t : ) : filter.tendsto (λ (x : ), (1 + t/x)^x) filter.at_top (𝓝 (real.exp t)) :=
begin

end

Adam Topaz (Jun 18 2021 at 00:28):

I don't know the answer, but it would be nice to have the more general form exp(t)=limx(1+t/x)x\exp(t) = \lim_x (1+t/x)^x.

Bhavik Mehta (Jun 18 2021 at 00:30):

Good point - I've amended the original post

Anatole Dedecker (Jun 18 2021 at 00:42):

I haven't done an extensive check but I would be surprised if those existed, since I've spent quite a lot of times using files where it could have been and have never seen it

Bhavik Mehta (Jun 18 2021 at 03:07):

I've got a roughly 50 line proof but I'm sure someone else will find a much nicer one!

Eric Rodriguez (Jun 18 2021 at 08:49):

Something similar is proved here: https://github.com/leanprover-community/mathlib/blob/700c4cb6f6d85b589d76f7e88fa9e26c74689149/src/data/equiv/derangements/exponential.lean (this is part of the derangements PR, fwiw)

The issue is that exp is defined using cau_seq, so it needs converting

Mario Carneiro (Jun 18 2021 at 14:06):

You shouldn't need to inspect the definition of exp here. real.pow is defined using exp, so you only need something like the derivative of log at 1

Sebastien Gouezel (Jun 18 2021 at 14:14):

Yes, we have everything we need on the exponential now, so its definition through cau_seq should never ever be used anymore.

Bhavik Mehta (Jun 18 2021 at 19:25):

Yup, Mario's suggestion is almost exactly what I did:

Bhavik Mehta (Jul 09 2021 at 17:41):

PR'd as #8243

Yaël Dillies (Jul 09 2021 at 18:13):

"The Fact"


Last updated: Dec 20 2023 at 11:08 UTC