My Math Forum Quartic equations - interesting problem

 Algebra Pre-Algebra and Basic Algebra Math Forum

 March 19th, 2018, 12:35 PM #1 Newbie   Joined: Oct 2013 Posts: 20 Thanks: 0 Quartic equations - interesting problem I can solve all square and cubic equations with complex coefficients. For quartic I use method from book "Solution of cubic and quartic equations" - S.Neumark. For $\displaystyle Ax^4+Bx^3+Cx^2+Dx+E$ In this method we first solving auxiliary cubic equation: $\displaystyle 1 -2Cx^2+ (C^2 + BD - 4AE)x -(BCD - BBE - ADD)$ next we choose y = any(?) complex root from three roots, Next compute: $\displaystyle den = \sqrt(B^2-4Ay)$ $\displaystyle G =\frac{B + den}{2}$ $\displaystyle g =\frac{B - den}{2}$ $\displaystyle H = \frac{C - y}{2} + \frac{B * (C - y) - 2 * A * D}{2 * den}$ $\displaystyle H = \frac{C - y}{2} - \frac{B * (C - y) - 2 * A * D}{2 * den}$ and 4 roots will be just two + two roots of two square equation: For $\displaystyle Ax^2+Gx+H$ and $\displaystyle Ax^2+gx+h$ only double square equations (B==D==0) need other simpler method - compute roots of square roots. Is OK, but 100/million random equations (random coefficient from -2 to 2 step 0.25) is bad. Bad if chosen cubic root is special: if real or imaginary part is zero, or first or second is 0.5 or 0.25. My test: Code: -0.50, -0.50 -0.50, 0.00 -0.50, 0.25 -0.50, 0.50 -0.25, -0.25 -0.25, -0.50 -0.25, 0.00 -0.25, 0.25 0.00, -0.25 0.00, -0.50 0.00, -1.00 0.00, 0.25 0.00, 0.50 0.25, -0.25 0.25, -0.50 0.25, 0.00 0.25, 0.25 0.25, 0.50 0.50, -0.25 0.50, 0.00 0.50, 0.50 1.00, 0.00 Is know this behavior? Is possible two or even three cubic root bad?
 March 19th, 2018, 01:13 PM #2 Global Moderator   Joined: May 2007 Posts: 6,770 Thanks: 700 It is not clear what parameters you are varying. In any case, from past experience, computer calculations can at times give weird results due to rounding effects. For 4th degree or higher polynomial equations I found that Sturm sequence approach works much better for real solutions.
 March 19th, 2018, 03:32 PM #3 Newbie   Joined: Oct 2013 Posts: 20 Thanks: 0 I found solutions, more important than roots $\displaystyle y_i$ is sum $\displaystyle B^2-4Ay_i$. If is near zero - it is bad. I must search root from cubic equation, which $\displaystyle B^2-4Ay_i$ has maximal modulus. If all are near zero - quartic equation has one quadruple solution. Problem is solved, but error propagation makes big errors in quadratic equations.

 Tags equations, interesting, problem, quartic

 Thread Tools Display Modes Linear Mode

 Similar Threads Thread Thread Starter Forum Replies Last Post xps32 Number Theory 9 September 17th, 2017 08:15 AM haswell Algebra 1 June 8th, 2015 05:17 PM simon naw Algebra 4 January 2nd, 2012 11:45 AM kustrle Algebra 3 February 5th, 2011 02:44 AM the_innovator Applied Math 8 March 22nd, 2007 08:44 AM

 Contact - Home - Forums - Cryptocurrency Forum - Top