Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.soft-sys.math.mathematica > #13946

Re: cubic equation solver

From Bob Hanlon <hanlonr357@gmail.com>
Newsgroups comp.soft-sys.math.mathematica
Subject Re: cubic equation solver
Date 2013-03-28 21:12 +0000
Message-ID <kj2bo4$m56$1@smc.vnet.net> (permalink)
References <20130328080639.90E5769F4@smc.vnet.net>
Organization Time-Warner Telecom

Show all headers | View raw


eqn = x^3 + (Sqrt[6] + 2 Sqrt[3] + 2 Sqrt[2] - 9) x +
    2 Sqrt[3] - Sqrt[2] - 2 == 0;

sol = Solve[eqn, x] // Simplify

{{x -> (-2*3^(1/3)*(-9 + 2*Sqrt[2] + 2*Sqrt[3] +
                 Sqrt[6]) + 2^(1/3)*
              (9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                   I*Sqrt[6*(2331 - 626*Sqrt[2] -
                            648*Sqrt[3] - 132*Sqrt[6])])^(2/3))/
         (6^(2/3)*(9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                 I*Sqrt[6*(2331 - 626*Sqrt[2] - 648*Sqrt[3] -
                          132*Sqrt[6])])^(1/3))},
   {x -> (2*3^(1/3)*((-9 - 6*I) + (2 - 3*I)*Sqrt[2] +
                 (2 + 9*I)*Sqrt[3] + (1 - 2*I)*Sqrt[6]) +
            2^(1/3)*(-1 - I*Sqrt[3])*
              (9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                   I*Sqrt[6*(2331 - 626*Sqrt[2] -
                            648*Sqrt[3] - 132*Sqrt[6])])^(2/3))/
         (2*6^(2/3)*(9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                 I*Sqrt[6*(2331 - 626*Sqrt[2] - 648*Sqrt[3] -
                          132*Sqrt[6])])^(1/3))},
   {x -> (2*3^(1/3)*((-9 + 6*I) + (2 + 3*I)*Sqrt[2] +
                 (2 - 9*I)*Sqrt[3] + (1 + 2*I)*Sqrt[6]) +
            I*2^(1/3)*(I + Sqrt[3])*
              (9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                   I*Sqrt[6*(2331 - 626*Sqrt[2] -
                            648*Sqrt[3] - 132*Sqrt[6])])^(2/3))/
         (2*6^(2/3)*(9*(2 + Sqrt[2] - 2*Sqrt[3]) +
                 I*Sqrt[6*(2331 - 626*Sqrt[2] - 648*Sqrt[3] -
                          132*Sqrt[6])])^(1/3))}}

Although not in the simplest form, these are the correct roots

eqn /. sol // Simplify

{True, True, True}

To find simpler forms without using FullSimplify (which seems to go on
indefinitely)

sol2 = ({x -> RootApproximant[#[[-1, -1]]]} & /@ sol) // ToRadicals

{{x -> Sqrt[5 - 2*Sqrt[6]]}, {x -> 2 - Sqrt[3]},
   {x -> -2 + Sqrt[2]}}

Verifying that these are exact solutions

eqn /. sol2 // FullSimplify

{True, True, True}

I assume that these are the values in your first sentence which was
garbled (presumably you did not convert to InputForm before copying).


Bob Hanlon


On Thu, Mar 28, 2013 at 4:06 AM, Elim Qiu <elim.qiu@gmail.com> wrote:
> x^3 + (=E2=88=9A6 + 2=E2=88=9A3 + 2=E2=88=9A2 -9)x + 2=E2=88=9A3 -=E2=88=9A2 -2 = 0
> has exact roots =E2=88=9A2-2, =E2=88=9A3-=E2=88=9A2, 2-=E2=88=9A3
>
> But Mathematica says:
>
> Solve[x^3 + (Sqrt[6] + 2 Sqrt[3] + 2 Sqrt[2] - 9) x + 2 Sqrt[3] -
>    Sqrt[2] - 2 == 0, x]
>
> {{x -> (1/
>       2 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>         I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>             264 Sqrt[6])]))^(1/3)/3^(
>     2/3) - (-9 + 2 Sqrt[2] + 2 Sqrt[3] + Sqrt[
>      6])/(3/2 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>         I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>             264 Sqrt[6])]))^(
>     1/3)}, {x -> -(((1 + I Sqrt[3]) (1/
>         2 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>           I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>               264 Sqrt[6])]))^(1/3))/(
>      2 3^(2/3))) + ((1 - I Sqrt[3]) (-9 + 2 Sqrt[2] + 2 Sqrt[3] +
>        Sqrt[6]))/(
>     2^(2/3) (3 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>          I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>              264 Sqrt[6])]))^(
>      1/3))}, {x -> -(((1 - I Sqrt[3]) (1/
>         2 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>           I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>               264 Sqrt[6])]))^(1/3))/(
>      2 3^(2/3))) + ((1 + I Sqrt[3]) (-9 + 2 Sqrt[2] + 2 Sqrt[3] +
>        Sqrt[6]))/(
>     2^(2/3) (3 (18 + 9 Sqrt[2] - 18 Sqrt[3] +
>          I Sqrt[3 (4662 - 1252 Sqrt[2] - 1296 Sqrt[3] -
>              264 Sqrt[6])]))^(1/3))}}
>

Back to comp.soft-sys.math.mathematica | Previous | Next | Find similar | Unroll thread


Thread

Re: cubic equation solver Bob Hanlon <hanlonr357@gmail.com> - 2013-03-28 21:12 +0000

csiph-web