| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8
Comments:
<0> that is an audacious piece of mathematica programming
<0> it's not as clean as i like (for a fun exercise, rewrite the program without using Table)
<0> but i'm glad it does work
<1> Thank you very much.
<0> no worries
<0> i often need something to keep me occupied of an evening
<0> let alone i have other more important work i need to get on with
<0> it's quite a simple program, even though it looks scary
<1> Well, I've used Mathematica like 30' in my whole life... it would take me hours to find out how to do this.
<2> isn't this easier?
<0> it gets easy if you keep practice
<2> I mean, all you want to do is to extend one vector to a basis right
<0> there may be an easier method
<0> but my code is to the method i described
<2> completeBasisR5[v_] := Select[Map[Append[#, v] &, Map[Delete[IdentityMatrix[5], #] &, Range[5]]], Det[#] != 0 &][[1]]
<0> galois: that's another algorithm, i'm sure it'll work to
<0> o
<0> the augment and row reduce method is the one i was originally taught
<2> okay this is dumb, I should use ReplacePart
<0> i try to keep to APL type functions sometimes
<0> as a matter of elegance
<2> completeBasisR5[v_] := Select[Map[ReplacePart[IdentityMatrix[5], v, #] &, Range[5]], Det[#] != 0 &][[1]]
<0> enough mathematica for today ;)
<3> hey folks, can I compute the mean square error for a vector?
<3> two vectors
<4> malik
<4> oops
<4> nvm
<5> I currently have some data:
<5> [ -1]: p: 0.3333 count: 7226 ratio: 33.45%
<5> [ 0]: p: 0.3333 count: 7023 ratio: 32.51%
<5> [ 1]: p: 0.3333 count: 7351 ratio: 34.03%
<5> at the moment I'm doing a chi-square test which says "this data signficantly deviates from the expected distribution"
<5> what I want is a test to say "the deviation in these data correlates with the intended deviation"
<5> any idea, firstly if there is a test, and secondly, which test to use?
<5> ok, easier question:
<5> is anyone alive?
<0> i personally hate statistics, so don't look at me
<5> :/
<1> :_GRR. Schur decomposition is evil.
<6> hi. i want to prove that \sum_{k=1}^n \frac{2^{n-k}{k} = \sum_{k=1}^n \frac{(-1)^{k+1}}{k} \sum_{j=k}^n \binom{n}{k}
<6> err \sum_{k=1}^n \frac{2^{n-k}{k} = \sum_{k=1}^n \frac{(-1)^{k+1}}{k} \sum_{j=k}^n \binom{n}{j}
<0> your TeX is still broke
<0> you mean \frac{2^{n-k}}{k} presumably
<0> this is why i use \over
<6> \sum_{k=1}^n \frac{2^{n-k}}{k} = \sum_{k=1}^n \frac{(-1)^{k+1}}{k} \sum_{j=k}^n \binom{n}{j}
<6> yes
<7> \over is evil
<0> but anyway, my brain's switched off for the evening
<0> \over is *not* evil
<7> what is thbbts?
<4> some feature of people down under i guess
<8> heh heh
<9> helloo
<4> rah : i don't your question
<4> it seems to the chi quadrat test gives you the answer you are looking
<4> but you cannot tweak the test outcome on a given data set
<0> it's blowing a raspberry
<0> it's onomatopoeic
<4> which means what in plain emglish ?
<0> oh, go look it up, i'm not your dictionary :)
<4> bah
<5> _kmh_: by "chi quadrant test", do you mean "chi squared test"?
<4> yes
<4> sorry my typing is quite lousy today
<7> anyway \over is not legal LaTeX, and will go away when hell freezes over and LaTeX3 is released
<5> _kmh_: the chi squared test tells me that there is a deviation
<5> _kmh_: correct?
<4> imho the 2 phrases you've given are just H_0 and H_1 of the chi chi squared test
<5> yes, there's a little bit of semantic overlap between "intended" and "expected"
<5> I want to know if ""the deviation in these data correlates with the deviation implied by H_1"
<4> it tells you whether the data contradicts you hypothesis vector
<4> rah : no
<5> no?
<10> can a partial order be a dag?
<4> you need to read that as _does not contradict_ H_1
<4> if you want an implication you need to formulate it negatively (if possible) so that H_1 becomes H_0
<5> ok
<0> spark: ***ume that a cycle exists in the repn of the partial order and try and get a contr from the partial order properties
<7> spark - eg the empty set is both a poset and a dag
<5> _kmh_: so what test do I need to do to ask "does not deviate in the specific way implied by H_1"?
<0> llll: i guess the nontrivial examples are more interesting
<0> ;)
<4> you only have an "implication" for H_0, for H_1 you only get a is not in contradiction to
<0> or, the possibility of nontrivial examples
<5> _kmh_: where "specific way" is that one cl*** will have *more than expected by chance*
<7> well 0<1 is both
<7> any linear poset is both, etc
<0> yeah
<5> _kmh_: I'm a bit lost
<0> llll: oh, i've just noticed your earlier remark about \over
<5> _kmh_: I want to check whether one cl*** ("bin") in my result set has significantly more results in it than any other
<0> llll: it's perfectly legal TeX, and that's what matters
<7> depends if you are writing TeX or LaTeX
<0> LaTeX is awful
<7> but the syntax is fairly evil anyway
<7> hmm doesnt amstex redefine \over?
<0> i'll check, doubt it
<4> rah : you probably need completely different test/test statistic then
<5> _kmh_: yes
<0> llll: outputs look the same
<5> _kmh_: the question I asked at the start was: which test do I need? :)
<0> i could go check the style files but i couldn't be bothered
<4> rah : dunno
<7> i mean \show\over
<0> eh?
<7> just checking the output isnt enough
<7> unless you cehck all possible inputs
<5> :/
<0> llll: don't get me started, it's way too late in the evening
<7> * rah looks at _llll_
<5> _llll_: do you know which test I need?
<7> oops
<5> _llll_: I want to check whether one cl*** ("bin") in my result set has significantly more results in it than any other
<0> i believe llll was talking to me there
<7> i have no clue about stats
<5> really?
<7> really.
<5> you told me about the Kolmogorov-Smirnov Test
<4> hehe
<7> i did? *I* dont even know what that test is
<0> llll: i think you referred rah to the R stats package
<0> R probably implements some sort of test
<5> oh
<5> ah
<5> yes
<5> R
<5> hmm
<4> it does implement a lot i guess
<0> doesn't mean llll knows what the test is ;)
<7> haha, i see, yeah i know that R is recommended for stats, that's abotu all i know
<4> you could check the efnet for R^^n, ***uming he's well tempered he probably will tell you what to do
<8> heh
<8> rtfm? :P
<0> ok, i'm going to bed now
<8> nini
<11> lolo
<11> R^^n? tetration?
<8> ??
<12> Maybe you meant: . v
<8> blah
<13> @keal
<12> i want to invent white dye
<8> ?
<7> ?b52s
Return to
#math
or
Go to some related
logs:
ZDE nothing happens zend
Could not open buddy file
#python
ubuntu dreamweaver replacement
#linux
#math
Fatal error: Call to undefined function mysql_pconnect() in /usr/share/php/adodb
jquery rollover
suse10.1 wmv mplayer
#lgp
|
|