- 14 Feb 2019, 20:47
- Forum: Helpdesk
- Topic: Problems with tropical polytopes with rational vertices
- Replies:
**4** - Views:
**40**

Which version of polymake are you using? We just released version 3.3, and there it works. It is possible that in some older versions the types get messed up. Beware that the property POINTS takes a small object of type Matrix<TropicalNumber<Min>>. Object of type Rational should automatically conver...

- 07 Feb 2019, 13:59
- Forum: Helpdesk
- Topic: Strugles with VERTICES_IN_FACETS
- Replies:
**2** - Views:
**61**

My purpose is to compute a cone $c pointed at some vertex $v of a polytope $p, such that its rays are the affine hull of the facets of $p incident to $v, and such that $p itself is not contained in $c. Sorry, I do not understand what you want to compute. Do you want to compute the intersection of t...

- 23 Sep 2018, 17:47
- Forum: General Discussion
- Topic: Math StackExchange Q&A - infinite family of self-dual polytopes
- Replies:
**1** - Views:
**6515**

Thanks for letting us now (and mentioning polymake). I like the Sierpinksi pattern!

- 27 Aug 2018, 16:40
- Forum: Helpdesk
- Topic: Facets calculated on 2nd try (not on 1st try)
- Replies:
**1** - Views:
**7250**

I looked into this, but I cannot reproduce your error. Sorry. Please install a version of polymake on your system, cf. https://polymake.org/doku.php/download/start . The shell.polymake.org version has a number of limitations and is proof-of-concept only. For instance, there is an upper limit on the ...

- 25 Jul 2018, 22:57
- Forum: Helpdesk
- Topic: 2-codimensional tropical hypersurfaces
- Replies:
**1** - Views:
**1020**

Your input needs to be homogenized.

- 04 Jun 2018, 14:03
- Forum: Helpdesk
- Topic: ERROR: no matching overloaded instance of...
- Replies:
**3** - Views:
**1330**

I'm getting errors of the form ERROR: no matching overloaded instance of... a lot, and I'm not sure how to interpret them. To give a bit more context: polymake supports operator overloading and polymorphism. That is, the function to be called is not only determined by the name of the function but a...

- 26 Feb 2018, 16:51
- Forum: Helpdesk
- Topic: Discrepancy in computing facets of a mixed integer program
- Replies:
**29** - Views:
**5087**

OK, I did one test with your example.

So cplex seems to be right about the LP.

Code: Select all

```
polytope> $lp=lp2poly "/tmp/problem.lp";
```

Code: Select all

```
polytope> print $lp->LP->MINIMAL_VALUE;
2917698461538468882086/17303076923076959125
```

Code: Select all

```
polytope> print convert_to<Float>($lp->LP->MINIMAL_VALUE);
168.623099493198
```

- 26 Feb 2018, 16:46
- Forum: Helpdesk
- Topic: Discrepancy in computing facets of a mixed integer program
- Replies:
**29** - Views:
**5087**

In the meantime, assuming you have access to an MIP solver (and therefore an LP solver) at your end that can read in the files I provided in the OP, I would be really grateful if any of the polymake developers can verify if the discrepancy exists on 3.2. My OP is detailed enough to be able to repli...

- 02 Feb 2018, 10:12
- Forum: Helpdesk
- Topic: Error when using projection command
- Replies:
**3** - Views:
**1848**

Please recall that projecting a polytope down by one dimension might (asymptotically) square the number of facets. Loosing nine dimension this might result in an enormous number of facets. As a side comment: exploiting the reverse is the idea behind the so-called "extended formulations"; see, e.g., ...

- 23 Jan 2018, 20:42
- Forum: General Discussion
- Topic: Lattice points at the intersection of 3 hyperplanes
- Replies:
**10** - Views:
**4826**

If you replace what obviously needs to be replaced in $p = new Polytope(INEQUALITIES=>{non-negativity constraints}, EQUATIONS=>{three hyperplanes}); print $p->LATTICE_POINTS; then this should do the trick. Several caveats: (1) depending on your hyperplanes the result may be unbounded. (2) Deciding w...

Powered by phpBB® Forum Software © phpBB Limited