Determinant Probability Distribution
When a matrix is filled with random numbers, its determinant becomes a random variable. The determinant probability distribution answers: for a randomly generated square matrix, how is the value of det distributed — which values are likely, and how often is the matrix singular (det=0)?
The Idea
The determinant collapses a whole matrix into one number measuring the "signed area/volume" spanned by its rows. For a 2×2 matrix (acbd) it is ad−bc. If the entries are random, ad−bc takes different values with different probabilities, and listing those probabilities is the distribution.
A Concrete Example
Place the numbers 1,2,3,4 (each used once) into the four cells at random — all 4!=24 arrangements equally likely — and compute det=ad−bc. The main-diagonal pair {a,d} and anti-diagonal pair {b,c} are always complementary 2-element subsets of {1,2,3,4}, whose products are
{1,2}→2, {3,4}→12;{1,3}→3, {2,4}→8;{1,4}→4, {2,3}→6.
So det can only equal ±(12−2)=±10, ±(8−3)=±5, or ±(6−4)=±2. Each of these six values comes from exactly 4 of the 24 arrangements:
| det | −10 | −5 | −2 | 2 | 5 | 10 |
|---|
| Probability | 244 | 244 | 244 | 244 | 244 | 244 |
The distribution is symmetric about 0: swapping the two rows negates the determinant, pairing the arrangements one-to-one. Note also that det=0 is impossible here — with four distinct entries the two diagonal products can never be equal.
Discrete vs Continuous Entries
The shape depends entirely on how the entries are generated:
- Discrete entries (drawn from a finite set, as above) give a discrete distribution over finitely many determinant values. …