[docs] fix bounds in big-M examples of tips_and_tricks.jl

[odow]

Follow-up to #3720

@remi-garcia how is this?

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.42%. Comparing base (af30cde) to head (2e6afd9).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #3721   +/-   ##
=======================================
  Coverage   98.42%   98.42%           
=======================================
  Files          43       43           
  Lines        5825     5825           
=======================================
  Hits         5733     5733           
  Misses         92       92           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[remi-garcia]

It is better, it is consistent that way. But I am wondering why in

In addition, if the variables involved have finite domains, then JuMP can automatically reformulate an indicator into a mixed-integer program.

variables must have a finite domain and not just upper bounds in case of <= and lower bounds in case of >=.

I will try to have a look at the JuMP reformulation as I think that it is possible to have JuMP to automatically reformulate an indicator into a mixed-integer program whenever there is the right bound and not just with finite domains.

[odow]

I will try to have a look at the JuMP reformulation as I think that it is possible to have JuMP to automatically reformulate an indicator into a mixed-integer program whenever there is the right bound and not just with finite domains.

Yes, that's possible. I just went with the simplest thing.

The relevant bridge is https://github.com/jump-dev/MathOptInterface.jl/blob/master/src/Bridges/Constraint/bridges/indicator_to_milp.jl