Wolfram Mathematica

« first day (2274 days earlier) ← previous day next day → last day (2211 days later) »

12:19 AM

I'm looking for a compact symbolic representation of an outer product of systems specified by strings or symbols. I want to use ⊗ but I'm worried about being having that redefined. \[Cross] might work, but I also want a syntax like {name, opts}\[Cross]{name, opts} to work and \[Cross] will evaluate for that.

Any ideas?

^the proper term might even be a Kronecker product (per J.M.)

5 hours later…

kirma

5:47 AM

RegionDistance[RegionDifference[FullRegion@2, Rectangle[]], {x, y}] ... what on Earth is going on in the result with those Roots?

2 hours later…

J. M. needs help.

8:09 AM

@b3m2a1 ⊗ is a bit overloaded in math (it is used for the outer product and the Kronecker product, for instance). However, CircleTimes[] in Mathematica has no built-in meaning, so you're free to use it as a custom binary operator.

b3m2a1

@J.M.needshelp. my major fear is that some other package or some user code will unkindly give it DownValues and then the feature will break. I may end up going with NonCommutativeMultiply because that one is protected... but it's not exactly clear that it conveys the meaning I wanted.

For anyone who's ever wanted to spelunk cloud objects:

Options[getCloudObjectPath] =
  {
   "Username" -> None
   };
getCloudObjectPath[path_String, ops : OptionsPattern[]] :=

  URLBuild@Flatten@{
     Replace[OptionValue["Username"],
      {
       s_String?(StringLength[#] > 0 &) :>
        "user:" <> StringSplit[s, "@"][[1]],
       _ -> ""
       }
      ],
     URLParse[path, "Path"]
     };
Options[getCloudObjectDS] =
  Options[getCloudObjectPath];
getCloudObjectDS[cpath_String, ops : OptionsPattern[]] :=

  Quiet@Check[CloudObjects[cpath], $Failed] //

Try this for some fun: getCloudObjectDS["Username" -> "b3m2a1"]

Or getCloudObjectDS["Username" -> "ResourceSystem"]

Ah shit you're gonna need to stick a CloudObjects[cpath, ops] in there

J. M. needs help.

8:31 AM

@b3m2a1 That won't be your fault anymore, if you explicitly state in the docs that you'll be overloading CircleTimes[].

b3m2a1

@J.M.needshelp. This is true, but I'm not even overloading CircleTimes. I'm just using it as a symbolic wrapper and it's kinda hidden in the package. It'd be in a template file used by a private function in a subsubcontext :) Don't expect anyone to really drill down that far.

b3m2a1

9:16 AM

Did the previous spelunk a bit better here: mathematica.stackexchange.com/a/170803/38205

3 hours later…

Kiro

11:47 AM

I'm trying to carry out the symbolic calculation here math.stackexchange.com/a/1176388/245806 in Mathematica 11.0.1 but don't seem to get the correct result.

My code is

Sum[Binomial[n, i] p^i (1 - p)^(n - i) Binomial[n, m - i] p^(
m - i) (1 - p)^(n - m + i), {i, 0, m}] // FullSimplify

I get

((1 - p)^(-m + 2 n) (-p)^m Gamma[m - 2 n])/(Gamma[1 + m] Gamma[-2 n])

which is different from the expected

Binomial[2 n, m] ((1 - p)^(-m + 2 n)) ((-p)^m)

Oh, forgot to mention that in my case I take the two binomial distributions to have the same number of trials n, and I try to find how it is distributed that two random numbers drawn sum to m

Kiro

12:05 PM

Okay, looks like it has to do with assumptions and the like as seen here mathematica.stackexchange.com/questions/41506/…

5 hours later…