Hi, I'm very new to the Wolfram Desktop Option System. I am simply trying to export a Dataset stored in a variable a .wdx file format and I get the following output:

What am I doing wrong?

Been off this site for quite awhile. Is "Guess who it is" rm -rf??

@MikeHoneychurch Nope. "Guess who it is" is our long lost moderator J.M.

rm -f is now "The Toad"

@halirutan ok. thanks for helping me catch up :)

Just like in the "rules of the internet": All men are men, all women are men and all children are undercover FBI agents.

lol

@MikeHoneychurch Btw, I'm @SjoerdC.deVries :-)

hi Sjoerd :)

Hi all.

I need some help with code optimization and good programming practice and code design

Would anyone here be willing to review a bit of code and suggest improvements?

@zab It chats it works differently because here we are in the party zone where usually no questions are asked. But if you think a quick question doesn't hurt and someone might answer, then simply ask.

Never ask whether you can ask because no one will volunteer until you presented your problem.

Haha - yes - I did notice the party mood everyone is in.

@zab, if it's quite long and involved, I suggest asking on the main site.

snippets should be okay here tho.

Well, the quick question is that I have written an algorithm to do some real life forecasting. There are a lot of exceptions and case switching, and the code is getting rather messy. I need help with good code organization practice.

The code works, but is clunky, and hard to edit.

Where do I go for this kind of thing?

@zab Hard to tell, because such questions will not draw much attention even on the main-site.

Exactly

I sure hope you don't have nested Which[]s three levels deep…

I'm happy to pay a consulting fee, if anyone is willing.

@zab Is it one file? About how much lines of code are we speaking?

No, it's not THAT bad

One file of code, about 150 lines

2 data files

Interested?

@zab Is it confidential or can you just post it to hastebin.com?

Can I email it to you?

It's confidential but anonymous, so should be fine

@zab That it's out of question that you can post it here anyway, right?

Yes - too big to post here I'm afraid

Also, I'll need a quick gmail (or similar) chat to explain what is going on.

@zab It's 3:48 in the morning here and I'm just finishing some of my own work. So I'm not open for consulting right now :-)

:-)

Understood

Later?

Tomorrow?

@zab I guess for such tedious work, you will have to look for a real consultant.

haha - yes true

Well, would you be willing to take a look and decide if you can be bothered?

I think it is not as bad as I made it sound

@zab Yes, I would be willing to take a look.

I'll add lots of comments to explain what is going on

email address?

@zab Uncompress["1:eJxTTMoPChZhYGAoSCwpykzOdshIzMksKi1JzNNLSQUAgX0JmQ=="]

haha - nice!

Are you in Berlin by any chance?

@zab Nope, but I know a very nice consultant from Berlin ;-)

Kuba?

I'm in Australia, and have family in berlin. Was there earlier in the year.

@halirutan I just realized how to send private messages

@halirutan I haven't used this chat before.

@zab Not private messages, but I get notified when you ping my name like this.

@zab I know Kuba and he is with the consultant I know.

@halirutan Ah right. OK then. I let you get on with your work. I just sent you an email.

@halirutan Kuba has helped me with similar problems in the past. Good man. But he's busy this week.

@zab How come that you could photograph all those australian celebs? (if this is not too private for a public chat)

Not at all

I'm a photographer in my spare time

@zab I guessed that one, but isn't it hard to get dates with all those famous people (I don't know how famous they are)?

@halirutan :-) They are famous for good reason - not just famous for being famous.

@halirutan I find people like that are usually very generous and gracious

@halirutan People who are famous just for being famous are hard to convince to pose for me

@zab The first thing you should do is to replace that crappy Wikipedia photo of Victoria Thaine with one of yours.

But I guess this would allow people to use your photo like they want.

@halirutan I did a series of portraits of artists, writers, Nobel laureates, scientists etc

@halirutan Victoria is a good friend. She's not too bothered about her public photos :-)

@halirutan Also, I shoot analogue, so the photos are mostly printed by hand, in the darkroom, and most are not digitized for web

@halirutan most people I photograph get exhibition quality hand made photos

Wow. The chemicals aren't cheap these days, y'know…

@halirutan The scans are not very good. My web design skills are even worse :-)

@zab uhh.. I have to admit that I'm using the complete opposite..

@halirutan Oh - the chemicals are very cheap. Time is expensive.

@halirutan In the end, I think the camera and the medium are just tools. The content of the photo is the most important thing. Whatever works for you is best.

Even the silver? (Or are you using a different process for developing?)

When I find the time to make some photos I'm using a Sony Alpha 77 which is digital through and through

@halirutan I have seen some incredible war journalism shot with an iphone.

"The content of the photo is the most important thing." - can't argue with that… :)

@halirutan Yes - film and paper are a little expensive.

@halirutan but to me, the biggest cost is time. It takes a whole day to do a shoot, develop the film, and make some prints.

@zab Since you always take a lot of photos with a digital camera and you need some time for post-processing, time is the limiting factor too for me.

@halirutan ah yes - true.

@halirutan What do you photograph?

@zab Depends, but mostly my family. Often I like to try things with HDR (since I'm working in image processing) but I have to admit I didn't touch my cam for some months now.

@halirutan :-)

@halirutan I'll email you some other "private" work - also a series for exhibition

@halirutan NSFW ;-)

@zab Ahh.. the current imgur discussion :-)

@halirutan I don't know it :-)

@halirutan I'm at work :-)

@halirutan which reminds me - I'd best press on.

@halirutan nice meeting you and chatting

@zab Yep. I should sleep a bit.

@halirutan will send you some email tomorrow

@zab OK.

@halirutan well, in a few hours, so you will see it tomorrow :-)

Can someone take a look at this question

and confirm that the success of the NIntegrate call depends on the variable name:

but...

…holy crap.

@halirutan I seem to get the same for x or z

@chuy How about using chi?

same

(all of these produce error.warning messages)

now did you define ff with _?NumericQ , because I have not

@chuy No, without NumericQ! You did it right

hmmm

very strange then

@chuy Yep. For Sjoerd it's the same as for you..

I mean the error message is "expected", but using x should make it "work"

should NOT that is

what OS + Mathematica version are you using?

I tried this at work where I have OS X and 10.1

64bit?

@chuy Yes.

On my linux box here it fails too, no matter which variable I use

It's not "failing" per se, but rather spewing warning messages

I mean I still get the correct answer

is that the same for you?

@chuy Well, but I got different answers.

now that is a key difference

@chuy And here on my Linux box, I get another answer..

on your Linux box are the answers correct despite the warning messages?

so on OSX when using x you get no warnings, but the wrong result is seems

@chuy Can you tell me what your answer to NIntegrate[ff[x]*x, {x, 0, 1}] is?

Here on Linux I get -1.67247 - 22.6365 I with and without NumericQ

I get the same

btw I am on Windows

does this act strange for you? NSum[ff[x], {x, 0, 10}]?

blah of course it should never mind

@chuy Let me make a simple example where I can see whether results are correct..

I am just goign to try a simple sum approximation

Total@Table[x ff[x]/100, {x, 0, 1, 1/100}]

I suspect the result from NIntegrate is probably right

@chuy Yeah, it probably is correct (at least here on Linux). Here is the analytical comparison when the interpolating function is just y(x)=x:

data = Transpose[{#, #}] &[Range[0, 1, .01]];
ip = Interpolation[data];
int[\[Xi]_] := NIntegrate[ip[rr] Cos[rr \[Xi]], {rr, 0, 1}];
NIntegrate[int[x]*x, {x, 0, 1}]

Gives 0.219886 on my machine

here the analytical result:

Integrate[chi*Integrate[x*Cos[x chi], {x, 0, 1}], {chi, 0, 1}]
N[%]

(* Out[33]= 1 - EulerGamma - Cos[1] + CosIntegral[1] *)

(* Out[34]= 0.219886 *)

so on OSX do you get an error message for ff[x] (by itself)?

@chuy Have to test this later. I'm not in the office anymore :-)

I see...

still very curious indeed

Can I set a parent stylesheet for a notebook which is dynamic? Let's say something like:

where pathToStyleSheet[] is by default "Default.nb" but I can evaluate in this notebook

pathToStyleSheet[] = "OtherStylesheet.nb"

and this should be reflected in appearance.

@MikeHoneychurch @JohnFultz @Mr.Wizard @halirutan any ideas? :)

More of them...

I think I've run out of ideas - and actually sensitive alternatives on this theme of whole little circles forming stuff.

hey @Szabolcs, I was reading your answer in here (mathematica.stackexchange.com/questions/1770/…) and it is related with a problem that a friend of mine just faced. He was just complaining that either ParallelMapor ParallelTable wouldnt give the same result as Map and Table (respectively) simply because the ordering of the parallel functions was not the original.

@Szabolcs - according to him, the ordering was very different and likely dependent on the speed at which the cpus were finishing the computations.

Do you think this is possible?

If anyone has any input into this issue I also appreciate it

I have a silly question about the above plot. Its made with the Show command, combining a listplot and a normal plot. I dont'get why there is this arbitrary line around 6; what causes this to show up? Plotting them individually doesn't show the line

I'm also using Frame, FrameLabel, PlotRange, PlotLabel and BaseStyle

@user3183724, try Axes -> None.

@Guesswhoitis. Yeah, that does it!

Good to hear.

@Kuba I'm not sure this is easily possible. One of the reasons is that you have to remember, that options in the form lhs -> Dynamic[rhs] are unlikely to work out of the box. I'm not sure how WRI implemented it that this works for things like Plot, but I'm not optimistic that it now works for all functions.

Consider this simple example:

Series[f[x] Exp[x], {x, 0, 2}, Analytic -> True]
Series[f[x] Exp[x], {x, 0, 2}, Analytic -> False]

(give me a sec.. wife needs help with the cooking)

@SquareOne Hmm!

@SquareOne In about one WolframDay

Do we all praise The Great Stephen on that day?

Actually, one of the speaker checked his $Version during his talk, and I saw it was released 17th of June ... Maybe a beta though ?

@Kuba OK, an even simpler example:

n = 1;
Slider[Dynamic[n], {1, 10, 1}]

Plot[x, {x, 0, 1}, Background -> Dynamic[ColorData[1, n]]]

This works, but that here doesn't

Plot[x, {x, 0, 1}, PlotRange -> Dynamic[{{0, n}, Automatic}]]

@halirutan I don't need panacea :) But I got your point. I think I have an idea, will have to try if it works :P

@Kuba This is the first thing that needs to be cleared. As far as I understood it, dynamic options are specifically handled and work only in rare cases.

@Kuba Good, because now the steaks are ready :-)

@halirutan chat.stackexchange.com/transcript/message/11191187#11191187

@halirutan and enjoy your meal :)