Wolfram Mathematica

3:06 AM

@FdotFloss This might help:

list = Sort@RandomReal[{-10, 10}, 100];

Position[list, x_ /; 2 < x, 1, 1]

(* OR *)

nf = Nearest[list -> Automatic]
nf[2]

It will give you one of the values you want. Position will return {} for a number greater than all x.

user11948

5:33 AM

thanks
I am solving two equations using FDM
I used nested for loops for this
[ For[iter = 1; iter < 100, iter++,
For[i = nt; i >= 1, i--,
For[j = 2; j <= nx, j++,
q[[i - 1,
j]] = Beta*Rho[[i, j]] + (1 + Alpha*(1/(nt - 1)) -
v*(nx - 1)/(nt - 1))*q[[i, j]] +
v*(nx - 1)/(nt - 1)*q[[i, j - 1]];
Rho[[i - 1,
j]] = (1 - Beta)*Rho[[i, j]] + Alpha*(1/(nt - 1))*
q[[i, j]];
]
]
]

nx=nt=100

Hi
I need some help

Smurfton

And? (BTW, not actually knowing mathematica, I only have a general idea of what this does.)

user11948

i imposed initial conditions
but for some reason
For loop is not doing anything
I am getting initial solution only as a result

Smurfton

Actually, that can't possibly be all of your code. Beta and Rho are undeclared.

user11948

Beta is a constant
q and rho are matrices dimension 100X100

Smurfton

nt starts at 100?

user11948

5:40 AM

nx and nt are 100 always

Smurfton

Has q already been filled with values?

user11948

yes q is zero everywhere except q[[nt,1]]= q0
and rho[[nt,1]]= (alpha/beta)*q0

i initialized rho and q like this and then the first loop is for no of iterations I want

Smurfton

[
For[iter = 1; iter < 100, iter++,
For[i = nt; i >= 1, i--,
For[j = 2; j <= nx, j++,
q[[i - 1, j]] = Beta*Rho[[i, j]] + (1 + Alpha*(1/(nt - 1)) -
v*(nx - 1)/(nt - 1))*q[[i, j]] + v*(nx - 1)/(nt - 1)*q[[i, j - 1]];
Rho[[i - 1, j]] = (1 - Beta)*Rho[[i, j]] + Alpha*(1/(nt - 1)) * q[[i, j]];
]
]
]

so...

What is q0?

is that also 0?

user11948

no its a constat

eventually my matrices should get filled with non zero values depending on no of sweeps i did

Smurfton

And, all of rho was 0, right?

except [[nt,1]]

user11948

5:46 AM

yes

Smurfton

So, assuming it hasn't been changed later, Beta*Rho[[i, j]] is 0.

user11948

yes during first interation

iteration

Smurfton

If types exist in mathematica, then 1+ Alpha *(1/(nt-1) is 1. In integer division, 1/99 = 0.

user11948

no that cannot be the issue
it will wither take that as 1/99 or to get the decimal value I need to put N

Smurfton

okay, then,

(1 + Alpha*(1/(nt - 1)) - v*(nx - 1)/(nt - 1))*q[[i, j]] = 0 on first iteration.

no.

sorry.

(1 + Alpha*(1/(nt - 1)) - v*(nx - 1)/(nt - 1))*q[[i, j]] = (1 + Alpha*(1/99) - v)*q0 on first iteration.

user11948

5:55 AM

yes

Smurfton

Actually, I was right the first time. the first iteration is at i= 100, j = 2. It never gets to i = 100, j = 1 because j is incremented in every iteration.

so, q[[99,2]] = v * q0

user11948

yeah i changed j =1
but I dont know why this nested loop in not giving any output

Smurfton

By the way, I don't know if this is like all other matrices in programming, but the first cell is 0,0 and the last cell is 99,99

usually.

Do you find out when there are logic errors present, or does nothing happen?

If that isn't the case, then q[[100, j]] is always 0, except when j = 1.

user11948

no mathematica is different it starts from 1,1
Anyways I guess I have to use C programming only
my objective was to find the answer in terms of alpha and beta only so that i can play with them but this is getting very frustrating now

Smurfton

Sorry that I can't help. :(

BESW

6:06 AM

← 42 messages moved from RPG General Chat

kirma

7:39 AM

Hmm, Wolfram Technology Conference 2014 materials: library.wolfram.com/infocenter/search/…

Mr.Wizard

9:17 AM

Someone please review my proposed duplicates (see comments) to this question and vote if appropriate:

0

Q: Display trailing zeros in ticks

When I specify tick marks in ArrayPlot I want to keep the number of trailing zero's that I specify. One work around is to use a label for that tick mark (i.e. Ticks->{0.8,0.9,{1.0,"1.0"},1.1,1.2}), but is there a neater solution to save me doing this manually? For example if I generate a plot wi...

plotting

dionys

9:45 AM

@Mr.Wizard It would be nice if the plotting functions could take an option for a function that would be applied to the default Ticks generated by a plot.

kirma

10:21 AM

Insert Wolfram Language... oh, it's nowhere to be seen.

halirutan

12:14 PM

Can someone tell me why I cannot see the "Join your favorite rooms" button sometimes? Like now:

Mr.Wizard

@dionys I agree! Unfortunately with an incomplete AbsoluteOptions function we cannot code that ourselves, from what I recall at least.

@halirutan Sorry, I don't know. Frankly I haven't noticed if it is always there or not.

halirutan

@Mr.Wizard It has not always been like that. I only noticed it recently at times..

@Mr.Wizard Another think I always wonder. Is it easily possible to only watch the chat without my icon being here? Because I have often seen that Leonid arrives as soon as we speak about some of his topics, but he wasn't pinged. So is it possible to watch incognito without logging out from SE?

Kuba

12:36 PM

@halirutan Ask Leonid, let's try to summon him: "Lazy evaluation"

Mr.Wizard

@halirutan You can view the chat transcript without joining the room. I don't believe it continually refreshes however.

Kuba

@Mr.Wizard It is joke that basic things like grid, tableview, ticks are no bugged/not finished. But hey, lets release WL for next platform...

Yves Klett

12:55 PM

@halirutan you need to log out (depending on your browser).

halirutan

@Kuba I was just having fun with grid yesterday. I had several histograms and when I combined them using grid and exported a pdf, the frames of the single plots aligned perfectly. Although, when I exported the plots separately and combined them using LaTeX's subfigure, then the frames had different sizes (although I fixed ImageSize and AspectRatio)

@YvesKlett That's what I tried to prevent..

Yves Klett

For some reason I always have to log onto chat separately...

halirutan

@YvesKlett hmm, no, when I'm logged into the main page, I'm always logged in at the chat too.

Yves Klett

@halirutan I have to do a whole SE roundtrip. Might be due to several blockers/add-ons, but who can tell.

halirutan

I upgraded to Ubuntu 14 lately and it contains bugs that weren't there in 10.. annoying

Look at my chat notifications.

Yves Klett

12:58 PM

@halirutan eeek... that looks almost BSODish

@halirutan NVidia?

halirutan

@YvesKlett Maybe it's a new effect and I should love it.

@YvesKlett Yes.

@YvesKlett But I didn't change the hardware and in Ubuntu 10 it worked.

Yves Klett

well, if this is any consolation - my gfx setup shows some strange bugs and even crashes on Win7. Sometimes google maps give me a complete system hangup.

halirutan

@YvesKlett hmm...

OK, need to run. See you all later.

Sander

Does anyone have a suggestion how I can find out what limitations there are on systems of SDE's that are recognised by ItoPocess? Or any other limitations I am likely to experience when scaling up the number of SDE's? I am testing a range of working models I imported into MMA. The smaller test cases (<5 SDE's) work fine, the one with 10 SDE's fails with "is not a random process recognized by the system".

Yves Klett

1:14 PM

@halirutan see you. Oh, and you could try opening chat in an incognito tab.

Kuba

1:47 PM

@halirutan @YvesKlett and ohers, let's say I have very sensitive Joystick. I would love to ba able to update ControllerState["X Rotation"] in Dynamic cell once per second. But signal from Joystick prompts Dynamic to evaluate 20x faster :)

I mean, I want to prevent the call on any link. I don't want to check the state and send it somewhere each sec. I just want it to happen once per second. For many events related to my Joystick preemptive link is quite crowded.

Yves Klett

2:19 PM

@Kuba sorry, no clue - interesting issue, though.

@Kuba can you actively poll the device via serial port? That should be doable.

Kuba

@YvesKlett :) p.s. ScheduledTasks seems to give some hope. Not issues-free approach though.

Yves Klett

@Kuba ... but I sure could use that accelerator of yours for several of my Manipulates. I would not mind 20x faster :D

Kuba

@YvesKlett well, it is trying but usually FrontEnd can't catch up :)

@YvesKlett it seems you have a good mood today :)

Yves Klett

2:42 PM

@Kuba ;-)

rm -rf

@halirutan Join in from an incognito window... you're not here, but the chat is live.

@halirutan It only shows up when you enter your first room. You can try now by leaving all the rooms and joining a room. It doesn't show up when you join a second room. (The button comes back if you're only in 1 room again, but you'll have to refresh the page)

Yves Klett

@rm-rf who knows, he already might be lurking.

Meta

6:18 PM

0

Q: How do I cut/paste Mathematica output into a question?

Of all things, I can't find any instructions in the editing-help about how to post my Mathematica output! Indenting code with four spaces works fine: f = a[x] + c1*b[x] + c2*c[x]; g = Expand[f*f]; h = Distribute[Integrate[g, x]]; Solve[{D[h, c1] == 0, D[h, c2] == 0}, {c1, c2}] But the outpu...

discussion formatting

rm -rf

6:55 PM

Hehe, I tweeted my favorite prank function to the bot.

belisarius

7:34 PM

@rm-rf You around?

@MichaelE2 I knew I had tried the Distribute thing! :D

Michael E2

@belisarius I was just writing a comment - incomplete as yet - that I think Jerry could use indefinite integrals for the algebra piece. Plug in the interval later.

belisarius

@MichaelE2 The problem is that hes defining a normed space, so he should use the that Norm with care

Michael E2

@belisarius True. I guess he could define a bilinear form (for AngleBracket) so as not to worry about premature evaluation.

belisarius

@MichaelE2 Yeap I'm sure there must be a package out there for QM that implements those things in a better way

Michael E2

@belisarius Yes, I'm sure there is.

rm -rf

8:01 PM

@belisarius What's up?

belisarius

8:11 PM

@rm-rf I was thinking about making a placeholder question on meta to allow users posting interesting statistic info they may find/elaborate about the site. What do you think?

rm -rf

@belisarius I think there's a similar question by Verbeia, although that might have been related to the very early beta days.

If that's too old/irrelevant, feel free to start a new one.

belisarius

I wasn't here on beta days. Lemme find it. And thanks!

rm -rf

5

Q: Statistics of the site

Is there a way to visualize the time-dependent statistics of our site, like how the requirements for coming out of beta changed so far? The small box to the right (Site Stats) and the Area 51 stats give only snapshots, though it would be nice to see the trends in a more straightforward way. Are w...

discussion feature-request

belisarius

Oh, I saw that one

But IDT it's the same

anyway, I'll read it carefulloy

rm -rf

8:32 PM

@belisarius This is the one I actually had in mind:

19

Q: How are we doing? (Post your favorite stats!)

By popular demand, here is a place for people to post regularly-updated metrics of traffic and other indicators of how Mathematica.SE is developing. Please post sets of analysis from different sources as (community wiki) separate answers.