Solve
for all real .
Write the quartic in standard form. Reordering by degree,
There are no or terms, but that does not make the equation solvable by substitution — the odd term rules out a biquadratic reduction.
Resist the temptation to drop the quartic term. Because looks negligible it is tempting to solve and quote . That is wrong: at the quartic term equals
which is five times the constant , not a small correction. A small coefficient is not a small term once is large — the crossover here is around , well below the naive estimate.
Bracket the roots by sampling. Evaluating :
So a root lies in . Sweeping across finds exactly two sign changes: one near and one near — the quartic has two real roots and two complex ones.
Bisect each bracket. Halving repeatedly until :
Newton's method from converges to the same positive value in three iterations, using .
Substitute both back. At : the quartic term is and the linear term is , summing with to ✓. At : the quartic term is and the linear term is , again cancelling the constant ✓. Note how both terms are comparable in size at the true roots — the opposite of the assumption that produced .
Need to solve a different problem like this? Open the solver →