@Pickett Works for me, too. Kinda funny DV, the answer being so old... +1

The the new Sequence family functions sounds great

SequenceCases SequenceCounts and SequencePosition

in V10.1

@SjoerdC.deVries Thanks for trying.

The could not delete error message was what I expected since I saw it the last time I tried it on a colleagues computer. It's a Windows problem.

The "Image (PP)" is one thing I have never looked at and never touched since it does not work on Linux. If the first click error persists for other users, I will look at it.

The ` Get["http://goo.gl/NaH6rM"]` error is very strange and I have it too sometimes. I tracked it yesterday (by using a table and calling it over and over again) and for me, it seems to be related to the waking the computer after hibernate. The URL itself should always work and it simply points to the latest version of the decode.m in the repository. I have created the short URL with Google and my login, so I always see it in my "short URL list at Google".

Very strange is, that Import seems to always work as it did for you. Therefore, we probably should replace Get with Import. The latter has the advantage to work in M8 too (with online resources like ours). If the palette has no other major problems on V8, then it is definitely the way to go.

@SjoerdC.deVries Regarding the "upload confirm": My small progress indicator work for you?

It is directly beneath the upload buttons and when it vanishes, the upload is done.

See here:

@Murta For example, this: reference.wolfram.com/language/ref/SequenceCases.html

Oh, seems that you've found them :)

@MichaelE2 Thank you

@halirutan Yes, it works. But I was thinking of a manual confirmation step, just as with images, so that you can see that you upload the intended cell/notebook.

@SjoerdC.deVries Actually, you don't get confirmation there. You only see the preview of the image and while it is uploading, the whole front-end hangs.

:-)

Since uploads of code will always look like the fox aehm wolf, I thought I skip this and since the uploading time can be quite large, I though that I don't use Szabolcs function and instead do it with a progress bar.

@halirutan I thought you could flash or highlight the cell/notebook and then the user confirms that you can go ahead or not. When I was talking about confirmation I was referring to the user side not the mma side.

For instance, for the image uploader I sometimes see in the preview that I have included the cell label, so I cancel the upload and select again.

@SjoerdC.deVries Yes, I see your point, but with cells or notebooks that can't happen. You select a cell or a many cells and you get them encoded.

@Murta they seem to be the most useful additions so far.

There is no rasterization step included that can go wrong.

True. The image uploaded sometimes does some unwanted line breaking.

And what do we do if the user selects many cells that are spread though the notebook? Or even worse, what do we show when a whole notebook is uploaded?

@SjoerdC.deVries Right, that's what I mean and this is why the preview for images is a good thing.

@halirutan to quote JFK: we do these things not because they are easy, but because they are hard. :)

@SjoerdC.deVries Hehe.. Well, I guess he was not talking about user experience :-)

@murta it's a pity that the older SequenceAlignment now appears to be a bit the odd man out as far as naming is concerned.

What I would dislike is another confirmation click if it is not necessary.

@SjoerdC.deVries Nice, this would be the very first function to make all my colleagues from bio-informatics trash Mathematica.

They should include another nice function SequenceAnalysis :-)

@halirutan ;) But seriously, I have been really missing SequenceCases and SequencePosition functionality. Their absence felt awkward. The former could be done with Replace* but not very elegantly.

@Mr.Wizard I usually vote but if I just remember a closely related question and don't have spare time to check if it is really dupe I'm just leaving a link.

the SE uploader palette has been missing it's buttons for a while for me now. is this a known problem? do I need to reinstall it?

@MartinBüttner yes, reinstall it according to instructions here

thank you

@halirutan I'm still worried about safety issues. Users should probably use the decode function from your package installed in their system and not using the google short URL. Otherwise it would be easy for someone to use a malicious decoder disguised as yours with a similar looking short URL (who remembers those cryptic URLs?).

Where can I get that marvellous uploader? I'm not quite certain if search results I see are fresh.

@Szabolcs "Sometimes no palette buttons show at all, just the Mathematica.SE logo on the left. If you can reproduce this, please contact me." I can, but I don't know what caused it.

@kirma what search results are you referring to?

I think this is an interesting question and I would like to see it edited into something the community deems worthy of reopening if at all possible:

@SjoerdC.deVries If it turns out that the rest does indeed work as we hoped it would, than I see absolutely no problem in creating a sub-domain on my page that links to the decoder. How about

decode.halirutan.de

?

@halirutan OK, if you can guarantee it'll work forever... or at least as long as either SE or imgur

otherwise we'd have the same problems as we have now with the documentation links breaking

@SjoerdC.deVries Well, as long as I pay my domain it will.

But that's a good point.

@SjoerdC.deVries Basically we have two choices: A cryptic URL like goo.gl or even on pastebin or a very long URL which speaks for itself.

How about a brute-force attack on google until we finally got

goo.gl/decode

@halirutan Third option would be to use your function name, with a footnote where to find the package. It requires a user action only once. Once installed, you can safely evaluate the function + argument that you copy from the post

It's a small inconvenience, but a big step in safety

@halirutan Even better perhaps if we could paste the PNG url in an inputfield of the palette and decode from there. That would centralize everything and make it more difficult to subvert the process.

@Mr.Wizard I agree that the question itself is interesting, but I feel that it is not really on-topic here. Shape-from-n-views is a research topic that has been the focus of deep and wide research in the areras of machine and human vision (I've studied some of this in the early 90's). It is very deep mathematically and although I think we might contribute a tiny but, I doubt we should.

but->bit

@SjoerdC.deVries Thanks for replying. So you feel there are no tools in Mathematica that would make it particularly suited to the problem? I don't expect anyone to push back the frontiers of research but I thought it was reasonably within scope. If the problem is to find any shape that satisfies the three silhouettes is that any more approachable?

@SjoerdC.deVries I agree! We don't need to use extra memory/lines of code with partition to do this kind of data manipulation now.

@Mr.Wizard But we already have answers on that, right? And one of the results was that the problem is not sufficiently constrained. I do believe Mathematica would make a fine tool for studying problems like these, but do we necessarily need this on the site? It's more about general algorithmic design than it is about Mathematica itself.

@Mr.Wizard But if you feel strongly about this I will reverse my vote of course

@Murta A strange omission. It's almost functionality you'd expect in V1.0

Hey guys

Im trying to show that there are only 18 prime numbers in the first 500 fibonacci numbers

My commands are list = Fibonacci /@ Range[500]

Length[Select[list, PrimeQ]]

However when I enter these commands first the table is shown with the first 500 numbers and then the number 18... how can i hide the table?

I guess the semicolon would work?

@Nickolas Yes, you want

list = Fibonacci /@ Range[500];
Length[Select[list, PrimeQ]]

Or

Count[Fibonacci[Range[500]], _?PrimeQ]

Or

Count[PrimeQ[Fibonacci[Range[500]]], True]

@SjoerdC.deVries I'm not actually certain of that any more...

@kirma I am working on an updated version of the SE uploader which has now the possibility to share cells and even notebooks.

That's what I'm discussing with Sjoerd all the time

@halirutan Ah, I was wondering if that functionality is already available. :)

@kirma It is, as beta testing thingy

@halirutan I'll gladly wait until it isn't, since I'm not really in dire need of all that...

@kirma No problem. It runs quite stable. I'm testing it currently on OSX and I have changed the banner logo again.

Hi all, I'm still on Mathematica 9 and I noticed that Mathematica 10 has a nice "ArrayResample" function. I tried creating one of my own in Mathematica 9 but it's horribly slow. I was wondering if it's OK to ask for a better way to do the ArrayResample in Mathematica 9, given that Mathematica 10 has the functionality built in.

@Guillochon Of course this is OK. Don't forget to include your own code and make clear, that you are aware that this is built-in in version 10 but you need it for version 9.

Many V9 users will thank you for that.

Resultant::poly2: -1+MeasureDumpR$230264^2+x^2+2 MeasureDumpR$230264 x Cos[IntegrateImproperDumpnewx] is not a well-formed polynomial in IntegrateImproperDumpnewx.... I guess I managed to break something again.

Hi everyone, i need to finish some code fast for my thesis but i'm having some problems and last night i even got sick . Is there anyone that can help me fix/improve it for a small fee? It's related to differential equation and root finding.

I don't see Dataset improvements in V10.1. No better memory handler for tabular data. I hope I'm wrong.

@Murta docs are still at 10.0. Do you have the beta?

Anyone?