| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12
Comments:
<0> in ranges, [ is inclusive, and ( is exclusive, right?
<1> So is a relation just a binary predicate?
<2> treitter - yes
<0> Catfive: thanks :)
<2> ihope - a relation between sets A and B is a subset of A x B
<1> Sounds like a binary predicate.
<3> ihope: yes, but on the level of sets rather than the level of logic
<3> I suppose one can translate between the two.
<3> P(x,y) if and only if (x,y) is an element of the relation.
<1> Yep.
<4> wut
<1> ?
<4> freenode gives me weird error messages
<1> Try ##help-unofficial.
<4> i don't want help
<5> hello
<4> hi@u
<4> are you having problems with the Z transform?
<5> does anyone have some xp with mathematical logic?
<4> depends how deep you are in it
<5> fraxtal, i read the elementary book. the first chapter is called "propositional algebra"...
<4> word
<4> what's yo problem
<5> fraxtal, it is general in nature. i cant seem to find some info on prop algebra on net. i mean it is not widely recognized as, for exemple, propositional calculus...
<4> well
<1> Are you sure it's not the same, then?
<4> i've never really heard it called propositional algebra
<5> http://www.gap-system.org/~history/Biographies/Church.html
<5> Haskell Curry in a review of the 1944 work writes:-
<5> It is written with the meticulous precision which characterizes the author's work generally. ... The subject matter is more or less cl***ical, namely, the propositional algebra and the functional calculus of first order, to which is added a chapter summarizing without proofs certain features of functional calculi of higher order.
<4> http://mathworld.wolfram.com/PropositionalCalculus.html is this not what you want to know
<5> no, this one is different. prop alg is concerned with semantics. propositions are interpreted as being false or true, and V => etc are interepreted as operations that take propositions, their falsity or truth, as their operands
<4> Maybe you're looking for First-Order Logic
<5> maybe
<4> try it
<5> let me check with wiki
<5> no it is not
<6> Boole treated truth values and logical operators as algebraic system
<7> hi, y=x^2-4x+3, y=mx+c, m=2x-4, simultaneous equasions. How do I get an expression from them for c in terms of m ? (from maxima)
<1> floundering: that's three variables and three equations?
<8> egad, into a tough M programing quackmire
<9> just realized that, adaptive sampling algorithms must also have a maxRecursion limit. (or something similar)
<9> otherwise, for some functions, it kept on computing almost forever
<9> i.e. kept on sampling.
<9> i spent a hour and added the maxRcursion functionality, but now i have to think about a design issue.
<9> basically, i'm writing a program that plots complex functions or 2D transformations in various ways.
<9> i.e. my modular surface, by riemann sphere, or simply transform any images in the plane, given a function f.
<9> first of all, by it's very nature, there must be a maxLength option. i.e. one must break the images so than neighboring points has distances less than maxLength. (only then, apply f)
<9> because otherwise, the function will only transform the extremities of the graphics primitives.
<9> for example, suppose you have a line from (0,0) to (0,1), and your transformation function is the z^2. (z being complex)
<9> if you don't have maxLength in your program, then you ends up with a post image that is another line, with extremities at (0,0) and i^2. Which is wrong.
<9> so, one needs to break the any preimage into many small polygonal pieces. That's where maxLength option came from.
<9> am i making sense? anyone?
<10> not a clue, sorrry
<11> which is a maximum step size
<9> Kahdloc: umm... if you are curious, i'll explain.
<11> for walking along a line (or curve)
<9> int-e: yeah, kinda.
<9> yes.
<10> xahlee, what's what you're doing called?
<9> Kahdloc: writing a program that plots 2d transformation functions with adaptive sampling feature.
<9> Kahdloc: say, a program for complex analysis.
<9> or linear transformation, affine transformation, ... etc.
<10> hmm...i've got a lot of reading to do :/
<10> xahlee, thanks for telling me. :)
<9> so far, for some functions such as inversion or complex inversion e.g. 1/z, the stepSize is not gonna do much goo because some parts blow up.
<9> Kahdloc: it just sounds more complex than it is. (i think a lot math is like that)
<10> yes.
<10> math is relatively simple. you just need to know the vocabulary.
<10> which i'm still working on :)
<9> so, then i implemented adaptive sampling. And today found a bug which tells me that by the very nature of adaptive sampling, one must also have a maxRecursion option
<10> what do you mean by 'maxRecursion'? how many times the function recurses?
<9> otherwise if i set maxDistance=0.2, and use function 1/z, it's gonna compute almost forever if my preimage is near the Origin.
<9> Kahdloc: yes.
<10> i suppose that'd make sense.
<10> have to limit the accuracy, or you'd never finish the graph.
<9> exactly.
<10> :)
<7> ihope: I am trying to find a tangent to two curves, I think thos equasions might be simultaneous at one intersection, if I can do that at both intersections I should have two equasions for m and c...
<1> Well, the system {y=x^2-4x+3, y=mx+c, m=2x-4} looks like it has a finite number of solutions.
<11> xahlee: an idea: if your path is z(t) and the transformation is f, pick maxDistance/||d(f(z(t))/dt|| as the step size. Some care is needed but then all the steps will have approximately the same length in screen space.
<12> c=y-mx, c=x^2-4x+3-mx, c=x^2-4x+3-(2x-4)x ?
<1> I'd eliminate m and y, then solve for x, then find the rest using that.
<11> xahlee: (Care is needed because that derivative may be 0 or close to 0. The idea also ***umes that your transformations and curves are smooth.)
<1> Actually, I think it has two solutions, but I'm not sure.
<10> how do you manage to remember all the odd quirks and rules of math?
<9> int-e: that's interesting. The program actually just map f to a set of points. So it actually doesn't do step size. The preimage can be any thing. e.g. a rect grid, polar grid, or post image of some other function...
<10> blunt repetition doesn't seem a good way to learn it :/
<12> oh, it was supposed to be in terms of m, not x :-)
<1> Kahdloc: who remember what?
<10> everybody
<10> this's #math after all :/
<1> Well... hmm.
<7> lde: yes, m and c are the same values at the other intersection but x is not
<11> xahlee: I'd still try to go from screen space to the preimages instead of the other way, wherever possible. That limits the amount of work per pixel
<9> int-e: oh i see. This is in Mathematica. :)
<9> int-e: yes, i know what you mean
<1> Kahdloc: well, the idea is to... well, um, yeah.
<1> Keep the quirks down. Learn either a little bit of everything, or everything about a little bit.
<1> Don't learn all of mathematics.
<10> hmm
<10> why not?
<1> That's too much to remember!
<10> ...i'm not so sure...
<7> ihope: yes thats my feeling, overall one positive and one negitive value for m, but only the negative one makes sense
<7> Ihope: I could be wrong there
<1> floundering: if only a negative value makes sense, then it's probably the negative value you want, aye?
<7> ihope; whuy-aye :)
<9> see http://xahlee.org/MathGraphicsGallery_dir/Transform2DPlot2_dir/transform2DPlot2.html
<9> the affairs of geometric transformation. BEAUTIFUL!
<13> beep
<14> BEEP BEEP
<15> hi all, anyone here use liboctave, or know how to call octave script from C++?
<16> http://a.ktug.lt/uploads/rowreduced.jpg
<16> this is not a row reduced matrix
<16> is it?
<17> it is
<16> thanks
<18> :o
<19> yeh
<19> it ****s
<18> oh really?
<18> hehe
<20> moreon : your nick is still cracking me up :)
<14> :)
<21> I am having issues with bayes theorem
<22> it's true i tell thee
<21> I have to show that P(A|B) = .1, I have P(B|A) = 0.75,, P(A|B) = .05 and P(B) = .6
<21> I have tried so many things but don't seem to get the right answer
<22> P(A|B) = P(AB)/P(B)
<22> is it?
<21> yes
<21> I get as far as P(A|B) = 0.75 * Pr(A)/ 0.6
<22> oh man i need paper
<22> I must admit i am confused
<22> I have to show that P(A|B) = .1, I have P(B|A) = 0.75,, P(A|B) = .05 and P(B) = .6
<22> you must show P(A|B) = 0.1 but you have P(A|B) = .05 ?
<23> P(A & B) = P(A|B) * P(B) = P(B|A) * P(A), divide by P(B) or P(A) and you get baye's
<21> nope
<21> P(A| not B) = .05
<21> I have P(A|B) = 0.75 * Pr(A)/ 0.6 and Pr(A) = 0.6Pr(A|B) + .2
<21> I am using bayes
<21> What I need is to get Pr(A) and then Pr(A|B)
<23> capital letters are generally used for random variables and vectors of all of it's possible outcomes. german ("small") letters generally denote individual outcomes. B = {b, ~b} for example. When saying for example P(A & B) = P(A|B) P(B) you mean to say any combination of A and B's outcomes (a,b) is the product of P(a|b) and P(b).
<23> Also recall that the probability of P(A = x) = Sum[ P(A = x | B = y) * P(B = y) ]. P(rain) = P(rain | cloudy) * P(cloudy) + P(rain | ~cloudy) * P(~cloudy) for example
<23> how far into science programs do you learn about contour integrals and what course names would indicate such?
<10> fuso, probably varies by location
<3> fuso: I'm not sure about science programs, but generally, later Calculus and Complex analysis courses.
<3> It wouldn't be unreasonable to talk about contour integrals in a second course on Calculus.
<23> Do you see any course that would cover it at http://www.matematik.lu.se/matematiklu/kurser/index.php?term=ht2006&sel=all&lang=en&print=no ? I would learn it on my own but I always feel like I'm missing some essential parts and if there is a course, I'd rather save it to concentrate on reading other things now
<24> Ugh. The integral of x/(x+2)^2 dx can be written as Log[x+2] + 2/(x+2) or Log[x+2] - x/(x+2). (+C)
Return to
#math
or
Go to some related
logs:
ubuntu drmsetbusid failed permission denied
#gentoo
debian aacraid etch
PACKAMN ON SUSE 10
LD_PRELOAD ubuntu
geronimo-specs missing dependency fedora core 5
#suse
#fedora
#css
CURLINFO_HTTP_CODE 302
|
|
