@WReach That works perfectly!! Thanks.

Is it just me or are Wolfram Community pages extremely slow to load? (And have been so more or less always...)

Dear friends, how to set V9 to be the default mathematica version while one have also v10 on board?

Win7. nothing works

@pickett I see you! help me :p

@OleksandrR. I presume your attempt has been using NelderMeadMinimizeDumpCompiledNelderMead rather than NelderMeadMinimize? I haven't tried the latter.

@Kuba I swear to God I as asleep :). Unfortunately I use OS X so I I can't help you out.

@Pickett ok, I believe you :) strange thing, win shows v9 is the default version but it switches to v10 anyway

@Kuba How do you run Mathematica?

@Öskå clicking on notebooks?

@Kuba Like going in the explorer and run it?

@Öskå yes

@Kuba And mma 9 is set as default?

yes

Yet 10 is launched? :D

even if I open v9 and click on notebook it opens it in v10

via explorer ofc

that's weird

I have to click each time open with v9 or open it though v9

Mhm, I'm sorry I can't help :o Never heard of such a problem. I guess it just takes the lastest version of mathematica.. Unix for the win is all I have to say :D

@Öskå I think it is good question for main but I'm not motivated to post it :p

main = SO?

@Öskå mathematica.se

@Kuba Well, that's most likely a Windows issue not a Mathematica one :)

@Öskå I don't have this problem with other things as movie players etc.

so I'd say mathematica issue narrowed to win OS

@Kuba Do you have two different versions of (for example) VLC installed?

@Öskå good question

I will post t here anyway :P

@Öskå MMA9 and MMA10 should be entirely separate applications with their own directories etc. It's more like Windows Media Player opening when VLC should open, and since that doesn't happen it could be a problem that is caused by the Mathematica installer which is on topic here.

Although Superuser.SE might know what this is about as well.

@Pickett Happy 10k btw :) OT = ?

@Öskå Thanks! I was using OT to say that it's on topic, only to realize that that's what people use for off topic as well...

@Pickett Sadly yes :P It's probably just Windows knowing that it has two different versions of the same Software and taking the lastest..

@Öskå I don't think so, I think it's this issue probably.

I don't think there is a concept of "versions" in Windows.

@Pickett I have no idea :) I stopped using Windows for a good reason ;o) Have to go eat lunch :) Laters :)

Have a nice meal. :)

@Kuba did you try to use Windows Systemsettings->Software->Default Proramms -> Filetype

@Karsten7. yes, it's v9 :)

@b.gatessucks hello, are you around?

anyone with v10 noticed problems with custom shortcuts?

@Szabolcs Hi

@Kuba what's the diagnostic?

@Kuba I am now

I am looking for a help on following question if any one has tried, any hint would be great, looking for belisarius. mathematica.stackexchange.com/questions/55495/…

@Mr. Wizard and @Simon Woods help request on following question: mathematica.stackexchange.com/questions/55495/…

@BeingHuman People are reluctant to help you because you have a long history of posting a question, linking to data, getting help and then removing the details from the question and the data. This makes the question useless for others and the answerers feel that their efforts are wasted.

@rm-rf that is a very serious allegations, I have all the data attached to my questions. Please check again. Let me know if missing something.

@BeingHuman Yes, it is a serious allegation — something for which I've contacted you before and something I stand by. Here's the most recent one:

People can see the full revision history to get a sense of what I'm referring to.

Please tell me @BeingHuman, how is version 6 in any way like version 3?

@rm-rf yes I am aware I am going to post it, but other than that I have pasted all other data for my questions. I feel you are being harsh on me.

@BeingHuman Your data is missing from this question and this one as well.

In any case, don't take it as criticism... you're not being punished or anything. Issues like this are discussed in here often and no one likes users that do that. I just wanted to let you know how people might feel.

@rm-rf for first one, yes, 2nd one it is in paste bin, please check, I will post other one paste bin right now, take it easy.

@rm-rf you sure like my questions and data a lot

I think they are interesting and we only have a few users here that are qualified enough to answer those questions. Your questions also require some effort to answer (which is good) and I only want that effort to be useful for others as well.

@rm-rf no disrespect, I will post it today, no disrespect, I owe to the community

@rm-rf actually for both cases: the data was already in paste link. I have modified question specific to that link.

@rm-rf except error question, I have updated the data link, so now you have all the data. for error question, I dont remember the file, I was trying my self, and got those errors, later on, I guess deleted the notebook. Any way, the conclusion was to understand polygons and points, which is well explained in the answer.

@rm-rf for the error question on acl comments, I have added the data in pastebin.

@acl please go ahead and edit it.

@BeingHuman done!

Suppose a Mathematica code that can be compile into a pure C program. Is it possible to run on computer without Mathematica?

@hwlau using Compile to produce the code?

I doubt it's possible.

It is possible using this sort of thing though:

SymbolicC/tutorial/Overview

(type this into the docs). Or online

@acl Is the only reason that it depends on the mathematica library

@hwlau I think so. But I only say this from looking briefly at generated code, I don't have anything deeper to say

@acl I have some complicated expression that generated by Mathematica, so I don't want to hand coding them again in C++

which is error prone and time consuming

@hwlau maybe this? reference.wolfram.com/language/CCodeGenerator/tutorial/…

@chuy it looks like the code that produces needs WolframRTL

@hwlau if they are long but otherwise not complicated, maybe CForm will help

It involves some special functions. The CForm may help a tiny bit.

@hwlau OK so you're hoping mma can export code to evaluate the special functions?

I am not sure it does this (I don't think so but could be wrong).

But if it outputs eg BesselJ(n,x) without defining it then you could just define it yourself, link to something etc

yes

even though it is no, I hope the other part are correct so that I can replace the special functino

@hwlau that I think can be done

@acl How?

@hwlau I mean, I think it does output something reasonable with CForm

@acl You are right!

why didn;t i try that

but I hope it can do it better

@hwlau I also just realised it now that you asked!

It is quite directly translated and not optimized.

I hope it can be better

but it is enough if it can directly generate exec codes

@hwlau what do you mean optimized?

Nevermind, it is just some evaluation order is faster in C.

@hwlau maybe this?

or this?

which is not quite important compared with the convience

but anyway shouldn't your compiler be able to take care of that? (I don't really know how good they are at this sort of thing)

@acl Compile can handle optimization of simple expression, not the long one

that why people recommend use newer compiler

or the intel one

but some long expression are obvious to my eye that they can be evaluated faster

particular those repeated expression that you dont want to evaluate twice

@hwlau OK, one of the packages I mention does try to optimize expressions. Don't remember which one and don't remember how well. But what they do is collect duplicate subexpressions, mostly, not more C-specific stuff if there is any

BTW, do you know whether mathematica can simplify expression with common part?

@hwlau right, I think one of those packages does it. Probably they're already included somewhere in mma now but I don't remember where

@acl which one?

@hwlau the package Optimize.m does this, from what I remember (the one I linked to above). Also something in mma does it, but I forget where precisely it is.

also look at second answer here (also first answer but the second seems ready-made)

@hwlau OK, here it is, built-in:

Experimental`OptimizeExpression[Sin[x^2 + y^2]/(x^2 + y^2)]

also take a look at its usage message, ?ExperimentalOptimizeExpression`

The output is ugly...

@hwlau add OptimizationSymbol -> a as an option

Where do you find these options? It doesn't work in my mathematica

These seems useful addon

Experimental`OptimizeExpression[Sin[x^2 + y^2]/(x^2 + y^2),
 OptimizationSymbol -> a]

I still remember I have to hand optimize the eigenvalue of the 4x4 matrix with thousands term

@acl Sor, I use it the wrong way

@hwlau last time I had to do this, I discovered the OptimizeExpressions thing

@acl I am not sure it will work for that case. It can be simplest because of the symmetry of the system and I have some expected results.

@hwlau Well, you can't expect mma to do physics for you!

The resulting given by mathematica is just some expansion, and it takes forever for it to simplify it

@acl The results are the same, just there are no meaning in the long expression :)

@hwlau I'm familiar with how it works. I find symbolic manipulation (with mathematica or anything else) pretty much useless for real work. I use it for numerics

(I mean, for my real work; obviously others find it useful for theirs)

@acl If you only need numerics, why are you still using mathematica :)

The OptimizeExpression is really optimize for real numerical evaluation. I believe they already use it in Compile already.

@hwlau because I like the language, I can write faster code in it than in python. And C doesn't buy you much if you are just diagonalising matrices.

and I can work orders of magnitude faster in mma than in C

@acl Same reason for me, but it is only few times faster than my C code.

(Well, it can be order of magnitude, but usually I have to wait for a day for it to finish the job)

@hwlau I should probably mention that I am not a good programmer in general, and a really (really) lousy C programmer

I actually find the combination works much better

but using two programming language at the same time is hard

@Szabolcs hi

had some time at work today to do a concept sketch for the flea-jacket-skates game

just a couple of lines of code in so far

selected = {1}; hosts = {{{5, 5}, 1}, {{-5, -1}, 2}}; Dynamic@
 Framed@Graphics[
   MapIndexed[
    EventHandler[{If[selected == #2, Green, Blue],
       Disk @@ #}, {"MouseClicked" :> (selected = #2)}] &, hosts],
   PlotRange -> 10]