As I wrote in the comment the second kernel uses a different pw-file playerpass. I moved this file and everything seems to work anyway.
But when I tried to start a second Mathematica, it told me that it weren't able to start a kernel due to licencing. This message popped up before the start-screen appeared were all the advertisement and the "New Notebook" buttons are.
@Szabolcs This whole screen looked really destroyed then.
Therefore, either the predictive interface is really started at the very beginning (and not with opening a notebook) or this second kernel is used for much more now.
Anyway, I was not able to kill the second kernel. The moment I killed the process, instantly another kernel was started.
I had had to delete my answer, because at first I accidently killed the wrong kernel (which obviously had the effect of killing predictions).
Also, another reason why that's NOT an SE ad is because it's the kind designed to lure you to click on it... SE ads are usually in-house ads, meaning links to other sites/questions (the real ads are only on SO, and are programming related)
I have a list of several hundred images, and I want to average the list to get a single image. Each image is 12bit TIFF with 2048*2048 pixels, and I am using the following to do the job:
result=Mean[Map[ImageData[Import[#]]&,imagelist]];
The problem is that the data becomes huge. I tried t...
@J.M. @acl Was trying to procrastinate putting the finishing touches on a paper and as a result was spending a lot of time in chat, being extraordinarily patient with some folks... nothing wrong with that, but a couple more days and he'd have attained "Tibetian Monk" status :P
@rm-rf actually, it's two papers finishing simultaneously. now, as the end of a paper approaches, the rate of progress decreases as 1/d with distance from end. but if it's more than one paper, then it's also 1/n, with n the number of papers. so overall now I am 1/2d with d a small number (which in addition does not change fast)
see, now I am even inventing theories of procrastination to avoid typing...
@acl You make progress only when it comes a full circle and you work on your paper to avoid doing the chain of N things that you tried to do to avoid working on your paper :)
Well, I generally don't edit the crappy and doomed to die anyway... only ones that might be salvaged (but that doesn't mean I won't close/delete a post that I edited)
@acl I say vote to close now... It's too localized anyway, and it can always be reopened if made substantial. Often people think "later" and it never happens, just adding to the unanswered list
btw, @acl @Szabolcs @halirutan @belisarius re: combating unanswered questions, here's a little known SE feature that you can use to remove the localized and unanswered questions — if a Q is >30 days and < 0 votes and has no answers, it will be deleted. So if you see a question where it fits the above criteria and is at 0 votes (e.g., OP never returned/responded), simply downvote it and it'll be gone the next day.
I didn't want to suggest this before or in the main thread because it could potentially be misused by anyone with just 125 rep. So just laying it out here for those that are active/well informed.
@acl the answers to the other question all concern differentiation, which is a completely different problem. Thus I can't see how they can be applicable here.
@OleksandrR. OK good point. To be honest, I originally thought the other question was also sort of off-topic (not really off-topic, but the answer really is "no it can't"). We just were excited and lots of people jumped in and wrote all sorts of clever answers.
But as you point out indeed they are specific to differentiation (which in fact wasn't even the point of the question!)
I love it when a new user immediately groks the usefulness of closing and removing useless posts:
Post probably useless for the Stack Exchange Community, especially with this title. That does not mean that you have to change the title. It's just that the post should be closed, I think. — andre6 mins ago
@OleksandrR. well, OK you're right, they're not OT. They are well-defined questions, on Mathematica, the actual answer to which is "No.". Then again, "Can Mathematica make coffee" would also be a valid question with this argument :)
Since the Tom Sawyer meta question it seems more questions of strictly limited usefulness are garnering TL votes. Probably a good thing although I do wonder how much of an effect it will have in the long run since with more 10k (or is it 20k now?) users every day, they will still be visible to a lot of people even when deleted and thus the dilution effect will remain
@OleksandrR. they're not actually visible to 10k users unless they have a link to it. So for all practical purposes, that question is hidden away. Within a 30 day period, they can see a link to it in the recently deleted posts in the tools, but not otherwise
Only mods can see deleted posts on user profiles and search with a deleted: 1 operator. I think it's for the best, as otherwise, there'll be a hell lot of clutter
@rm-rf oh, I see. I was only briefly able to see those posts before graduation so I didn't take notice of the criteria for it. Deleted posts being actually deleted is much better, of course.
@halirutan That's something that does not exist. You can only search for posts with X votes or more, not less. The way I do it is to sort by votes and jump to the last page and walk backwards :)
@OleksandrR. Since it came up, here's the current tally:
@Verbeia I assumed, that many live in different timezones than my own. Although I'm a night-owl and can therefore chat with the US guys/girls sometimes.
@OleksandrR. yes, it has decreased (as a % of total posts) these days... it used to be very high when we had an insanely large number of folks with instant delete vote privileges.
Incidentally, although I strongly agree with Leonid's sentiment regarding falling average question quality, I also think that the wish to remain indefinitely in an "age of innocence" is not completely realistic. We as a site have now entered adulthood and will have to learn how to deal with it.
After the performance problems we have run into with Lucene.NET we've decided to make a change, we're moving the network on to elasticsearch.
Here's where to get started: http://stackoverflow.com/search
What works:
All search operators should be in
Many changes below from the old search behav...
It's a different dynamic – mods only live inside the flag queue, don't have time to answer or edit posts. There is no sense of community; They don't hang around in chat to socialize or discuss stuff, etc. They more or less visit meta only when some hurt fellow posts a rant or some really interesting discussion is happening...
A lot of the flags on SO are easy to handle — i.e., they get so much filth, that just from the snippet shown in the mod panel, it's obvious that it's spam/crap, so they just blast away. On the other hand, they have very complex problems too — strategic downvoting, fraud, heavy sockpuppeting, vote rings, etc. that takes a lot more time to dig through and investigate
Sounds like quite an unpleasant job, overall. Going from being an enthusiastic user to someone who is so overwhelmed with crap that they don't have time to participate any more must be a shock.
@rm-rf I also notice that meta.SO has a rather different content to meta on other sites. It seems to be full of in-jokes and cliquey nonsense that make it hard to see the wood for the trees.
@OleksandrR. On the contrary, a lot of them seem to want it, despite knowing fully well what the job entails. I guess if you're a full time software developer, a strong SO cv and moderatorship can be very helpful.
@OleksandrR. meta.so was meant to be the meta site for the trilogy (SO/SF/SU), but then became a network wide meta and SO's support site. There's a lot of talk about cutting down the memes and stuff, but from what I understand, it was a lot worse 3-4 yrs ago (it was basically a meme fest)
@rm-rf maybe so. Does SO participation count for anything IRL, do you think? Personally I doubt if someone will be promoted to a management role based on their success at being an SO mod, but maybe things work differently to what I imagine
@OleksandrR. Oh yes, it most certainly does; but it varies from tag to tag. If you're active in C#/Java/Javascript/PHP/Ruby/Python/etc., you have a very good chance of getting a good job via careers.so (that's the money maker)
For mathematica/matlab/maple and similar smaller tags, not so much.
@Verbeia it's certainly impressive given that you're not actually in IT and have other rather important duties to attend to. :) But you have to admit, your SO/SE rep is not going to be considered an important qualification in your line of work, as I suppose is the case for most of us on this site... it's just a hobby. I was wondering if it goes further than that if you are actually working in an IT-related field, and apparently it does...
@OleksandrR. you're quite right the SE network rep has no bearing on my actual career. Actually someone has re-proposed Economics on Area 51, and I think it will fail again.
@Verbeia it seems to me that people who have something to say about economics usually just set up their own blog. That leaves those who have silly questions to ask and those who can't tell an economic question from a political one
@OleksandrR. exactly! I would never have wasted my time in the old Econ.SE site. Also, the subject just doesn't lend itself to SE style QA. There are so many people peddling obviously wrong stuff, but an amateur or an "Austrian school" type might accept that crazy answer anyway.
hi. wondering if anyone might offer some suggestions on a question I asked previously mathematica.stackexchange.com/questions/21927/… ? had thought that the offered duplicate question might help but since I have a Power in my expressions FullForm sadly won't work.
@rm-rf I suppose in the old style using ReplaceAll this was sort of automatic. I don't know if there's any alternative available with the post-6 methods.
@Verbeia I think in economics, more than in any other academic discipline AFAIK, there are so many opposed schools of thought with mutually contradictory views (most of which are also constantly at each other's throats) that trying for a single "correct" answer is more or less impossible... almost everything is subjective if you don't limit your scope to one particular school. The same problem seems to have arisen on Islam.SE, actually
@halirutan yes. it's tricky as really want to transform a FullForm from Power[x,Times[11,Power[y,z]]] to Power[r, Power[y, Plus[-1, z]]]; where one is taking a factor out of y^z as y*y^(z-1). would be great to automate this stuff.
@halirutan @bsdz The solutions on the dupe don't seem to be directly applicable in this case. I'll reopen it, but please rephrase the question to 1) Not ask about structural manipulations (you most definitely need a mathematical approach here) and 2) mention that the solutions in the other question don't work
@rm-rf I have this functionality implemented in the past, in packages CheckOptions and PackageOptionChecks - although I am not sure that what I did is what you are after. Basically, they auto-generate new function definitions, by parsing function's signatures
@halirutan actually that's not what I want to do, that's replacing the whole lhs with a rhs. I was showing the fullforms not of the LHSs I wanted to transform but the expression before and the expression after.
@halirutan sure. I want to do this x^(11 *y^z) - 1 /. { x^(11*y) -> r } and have Mma return r^(y^(z - 1)) - 1; I think @rm-f might have provided a solution although not sure if that would work on the other expressions I need to apply the same transform too.
@rm-rf A section "Changing option-protecting information from outside of the package" in that notebook contains a minimal stand-alone example of how to option-protect a single function. But OptionsPattern[] is not currently supported, only ___?OptionQ. Adding that support shouldn't be hard though.
@bsdz Michael gives a possible solution in the comment. As you see, you have to provide assumptions to the variables because (as I said) this transformation is not true in general.
@halirutan Furthermore, please note that this is not just a simple replacement. You need Simplify to get your z-1 expression. In more complex situations this might not work.
@bsdz Btw and since you are new: You did it perfectly well. Even when your question was closed as dup come to chat ask nicely for help and show that you made some efforts to find a solution. Anyone is willing to help you then.
@LeonidShifrin Yes, that package looks like it does what I intended. I'll probably go with my little function above (a modified version of it) since it's short and works well (and is internal to a package, not for general use), but I got a few good ideas from how you implemented it, which I will steal :)
@OleksandrR. Not the worst by a wide margin... I'm sure he got dinged for his comment: "Put some values for r and alpha in my original post and you will get the idea, that's basic math." when asked to make an MWE for his indecipherable question.
@acl I have difficulty deciding whether the one with the OP asking why his keygenned kernel didn't work is the worst, or if the award should be given to one of the several questions where the OP presents some unintelligible problem statement and expects all the work to be done for them
@acl Well, there was one that was basically: "I'm new to mma and was asked to make a complex simulation on the dynamics of interaction between organizations. Here are my parameters, how do I do it?"
@rm-rf oh, so I can see the user's name on the other question only because they edited it themselves, then, and so their name shows up in the revision history?
please don't post it again though... as much as their questions here are fun, and their personal site is googleable , it's not fair to them to post it publicly for laughs, especially when they did not make it public here.
@rm-rf further on options: say you have functions f and g that share many options but might have different defaults. You wish to call g inside f, but in this case the default options for g should be overridden by the defaults for f. Can you think of a better way to do that than to give all the options for g explicitly in the call using OptionValue[f, opt]?
Problem is that FilterRules only filters rules given explicitly, not as defaults for a function. OptionValue invokes "magic" to unify explicit options with defaults, but one can only extract individual option values that way, not a complete list of options