Re: Re: Trouble with a system of equations

*To*: mathgroup at smc.vnet.net*Subject*: [mg77666] Re: [mg77596] Re: Trouble with a system of equations*From*: Andrzej Kozlowski <akoz at mimuw.edu.pl>*Date*: Thu, 14 Jun 2007 06:15:53 -0400 (EDT)*References*: <f4lb44$fvv$1@smc.vnet.net> <200706131142.HAA07187@smc.vnet.net>

On 13 Jun 2007, at 20:42, Yaroslav Bulatov wrote: > On Jun 11, 10:31 pm, Ray Koopman <koop... at sfu.ca> wrote: >> On Mon, 11 Jun 2007 07:38:00 -0500 drmajor... at bigfoot.com wrote: >> >>> There's an unmatched bracket in >> >>>> m = Inverse[ Subsets[Times@@#]& /@ Tuples[{0,1},k]] ]. >> >>> and I haven't found a way (so far) to correct it so that the code >>> works. >> >>> Bobby >> >> Ah, the joys and perils of posting code without testing it first. >> Maybe someday I'll get Mathematica for my machine at home. >> >> Aside from the extra ], that must must have been teleported from >> LinearSolve[x,Log[p/(1-p)], where a ] is missing, the problem is >> that I simply copied the form of an example in the Subsets online >> documentation, with h changed to Times, not realizing that Times@@# >> would be evaluated before Subsets got to it. >> >> Here are two ways to get x: >> >> ReleaseHold@Subsets[Hold@Times@@#]& /@ Tuples[{0,1},k] >> >> or (preferably, I think) >> >> Times@@@Subsets@#& /@ Tuples[{0,1},k]. >> >> With[{k = 2}, Inverse[ Times@@@Subsets@#& /@ Tuples[{0,1},k] ]] >> >> {{ 1, 0, 0, 0}, >> {-1, 0, 1, 0}, >> {-1, 1, 0, 0}, >> { 1,-1,-1, 1}} >> >> With[{k = 3}, Inverse[ Times@@@Subsets@#& /@ Tuples[{0,1},k] ]] >> >> {{ 1, 0, 0, 0, 0, 0, 0, 0}, >> {-1, 0, 0, 0, 1, 0, 0, 0}, >> {-1, 0, 1, 0, 0, 0, 0, 0}, >> {-1, 1, 0, 0, 0, 0, 0, 0}, >> { 1, 0,-1, 0,-1, 0, 1, 0}, >> { 1,-1, 0, 0,-1, 1, 0, 0}, >> { 1,-1,-1, 1, 0, 0, 0, 0}, >> {-1, 1, 1,-1, 1,-1,-1, 1}} > > OK that works, thanks. One problem with the original system is that I > set up the estimation equations in terms of E[x_i y] (assuming uniform > distribution over x's), and not empirical odds ratios which apparently > make it much harder to solve (Andrzei's solution works fine for 2 > variables, but takes too long for 3) > > I think a variant of my method also works for three variables. I will soon sent you a notebook, after adding to it sme comments. Andrzej Kozlowski

**References**:**Re: Trouble with a system of equations***From:*Yaroslav Bulatov <yaroslavvb@gmail.com>