Discriminant

Deep Breakdown

What it tells me (Situation Awareness): The discriminant $\Delta = b^2 - 4ac$ is the part under the square root in the quadratic formula. Before solving for the actual roots, the discriminant tells you how many real solutions exist and what type they are — without any additional computation. In data work, it is useful when you need to quickly classify whether a quadratic model has viable (real) solutions, such as whether a break-even point exists at all.

Flowchart

Real-world use

Where it is used

Determining feasibility before solving: does a price-demand model have a break-even point? Does a trajectory model hit the ground? Does a signal model have real-valued resonance frequencies? The discriminant answers these questions without computing the full formula.

Problem it solves

Computing the full quadratic formula is wasted effort if the roots are complex (i.e., no real solution exists). The discriminant is a fast pre-check: compute it first, classify the situation, then decide whether to proceed.

How it solves the problem

The discriminant is the expression inside the square root: $\sqrt{b^2 - 4ac}$. If this is negative, the square root is imaginary, meaning no real $x$ satisfies the equation. If zero, the $\pm$ gives the same result: one repeated root. If positive, two distinct roots exist.

Example in a business

Revenue model: $R(p) = -p^2 + 50p$, fixed cost $C = 700$. Break-even equation: $-p^2 + 50p - 700 = 0$ (multiply by $-1$: $p^2 - 50p + 700 = 0$). Discriminant: $50^2 - 4(1)(700) = 2500 - 2800 = -300 < 0$. No real break-even price exists — revenue never reaches $700$. The discriminant alone flags this: no need to apply the full formula.

The formula, piece by piece

• $\Delta$ (delta): the Greek letter for "discriminant" — it "discriminates" between solution types
• $b^2$: the square of the linear coefficient
• $4ac$: four times the product of the leading coefficient and the constant
• $b^2 - 4ac$: their difference determines the geometry of the parabola relative to the $x$-axis Discriminant decision table: | $\Delta$ | Roots | Parabola behaviour | |-----------|-----------------|--------------------------------------| | $> 0$ | Two distinct real | Crosses $x$-axis at two points | | $= 0$ | One repeated real | Tangent to $x$-axis at vertex | | $< 0$ | No real roots | Entirely above or below $x$-axis |

The formula in words

"Compute $b$ squared minus four times $a$ times $c$. The sign of this number tells you how many real solutions the quadratic has." $\Delta > 0$: "The parabola crosses the $x$-axis — two real solutions." $\Delta = 0$: "The parabola just touches the $x$-axis — one solution (a double root)." $\Delta < 0$: "The parabola misses the $x$-axis entirely — no real solutions."

Critical Concepts

• Discriminant is not a root. It is a diagnostic value, not the solution itself.

• $\Delta = 0$ means a perfect square trinomial: $ax^2 + bx + c = a(x - r)^2$ for some $r = -b/(2a)$.

• Large $|\Delta|$ means roots are far apart. Small $\Delta > 0$ means roots are close together (near-tangent case, numerically sensitive).

• Sign of $a$ and sign of $\Delta$ together determine the parabola's shape: - $a > 0$, $\Delta < 0$: parabola opens up, always positive — no crossing.

• $a < 0$, $\Delta < 0$: parabola opens down, always negative — no crossing.

• In statistics: $\Delta$ relates to conditions for positive-definiteness in certain quadratic forms.

Step-by-step calculation (tiny example)

Step 1: Read off a, b, c from the equation Equation: $3x^2 - 7x + 2 = 0$. $a = 3$, $b = -7$, $c = 2$.

Step 2: Compute $b^2$ and $4ac$ separately $b^2 = (-7)^2 = 49$. $4ac = 4 \times 3 \times 2 = 24$.

Step 3: Subtract and classify $\Delta = 49 - 24 = 25 > 0$. Two distinct real roots exist. Full roots: $x = \frac{7 \pm \sqrt{25}}{6} = \frac{7 \pm 5}{6}$, giving $x_1 = 2$, $x_2 = \frac{1}{3}$.

Code examples

Beginner example (Python)

def classify_discriminant(a: float, b: float, c: float) -> dict:
    """Compute and classify the discriminant of ax^2 + bx + c = 0.

    Args:
        a: coefficient of x^2
        b: coefficient of x
        c: constant term

    Returns:
        Dictionary with discriminant value and classification
    """
    delta = b**2 - 4 * a * c
    if delta > 0:
        classification = "Two distinct real roots"
    elif delta == 0:
        classification = "One repeated real root"
    else:
        classification = "No real roots (complex only)"
    return {"delta": delta, "classification": classification}


# Test cases
test_cases = [
    (1, -5, 6),  # 2 real roots
    (1, -2, 1),  # 1 repeated root
    (1, 0, 1),  # no real roots
    (3, -7, 2),  # 2 real roots
]
for (a, b, c) in test_cases:
    result = classify_discriminant(a, b, c)
    print(f"{a}x^2 + {b}x + {c} = 0: Δ = {result['delta']:>5} -> {result['classification']}")

Beginner example (R)

classify_discriminant <- function(a, b, c) {
  #' Classify the discriminant of ax^2 + bx + c = 0.
  #' @return List with delta and classification
  delta <- b^2 - 4 * a * c
  classification <- if (delta > 0) {
    "Two distinct real roots"
  } else if (delta == 0) {
    "One repeated real root"
  } else {
    "No real roots (complex only)"
  }
  list(delta = delta, classification = classification)
}
test_cases <- list(c(1, -5, 6), c(1, -2, 1), c(1, 0, 1), c(3, -7, 2))
for (tc in test_cases) {
  result <- classify_discriminant(tc[1], tc[2], tc[3])
  cat(sprintf("%dx^2 + %dx + %d: D = %5g -> %s\n", tc[1], tc[2], tc[3], result$delta, result$classification))
}

Professional example (Python)

import numpy as np
import pandas as pd

# Scenario: feasibility screening for quadratic pricing models
# Before computing full roots, use discriminant to flag infeasible cases
np.random.seed(42)
pricing_models = pd.DataFrame({
    "product": ["A", "B", "C", "D", "E"],
    "a": [2.0, 1.0, -1.5, 3.0, 0.5],  # coefficient of p^2
    "b": [-80.0, -40.0, 60.0, -30.0, -20.0],  # coefficient of p
    "c": [600.0, 900.0, -500.0, 400.0, 250.0],  # constant (fixed cost)
})
pricing_models["discriminant"] = (
    pricing_models["b"]**2 - 4 * pricing_models["a"] * pricing_models["c"]
)


def classify_delta(delta: float) -> str:
    if delta > 0:
        return "Feasible (2 break-even prices)"
    elif delta == 0:
        return "Marginal (1 break-even price)"
    else:
        return "Infeasible (no break-even exists)"


pricing_models["feasibility"] = pricing_models["discriminant"].apply(classify_delta)
print("Break-Even Feasibility Screen using Discriminant")
print("Model: a*p^2 + b*p + c = 0 (find break-even price p)")
print()
print(pricing_models[["product", "a", "b", "c", "discriminant", "feasibility"]].to_string(index=False))
print("\nProducts with negative discriminant cannot break even at any price — flag for review.")

Professional example (R)

# Break-even feasibility: discriminant b² - 4ac
set.seed(42)
pricing_models <- data.frame( product = c("A", "B", "C", "D", "E"), a = c(2, 1, -1.5, 3, 0.5), b = c(-80, -40, 60, -30, -20), c = c(600, 900, -500, 400, 250)
)
pricing_models$discriminant <- with(pricing_models, b^2 - 4*a*c)
classify_delta <- function(delta) {
  if (delta > 0) "Feasible (2 break-even prices)"
  else if (delta == 0) "Marginal (1 break-even price)"
  else "Infeasible (no break-even exists)"
}
pricing_models$feasibility <- sapply(pricing_models$discriminant, classify_delta)
cat("Break-Even Feasibility Screen using Discriminant\n")
print(pricing_models[c("product", "a", "b", "c", "discriminant", "feasibility")])

Source: 1300 p. 21

#algebra#powers-roots-and-inequalities#equations-and-inequalities