Matrix Norm Calculator

Enter the four entries of your 2x2 matrix to find its 1-norm and infinity-norm.

How it is calculated

The matrix is written as rows [a b] and [c d]. Each entry is replaced by its absolute value, so signs never matter. The 1-norm adds the absolute values down each column and keeps the larger total: ‖A‖₁ = max(|a|+|c|, |b|+|d|). The infinity-norm adds across each row and keeps the larger total: ‖A‖∞ = max(|a|+|b|, |c|+|d|). The tool also reports the largest single absolute entry, max(|a|,|b|,|c|,|d|). All entries are plain numbers with no units, and results are rounded to two decimals. It does not compute the spectral (2-)norm or the Frobenius norm, which need squares and eigenvalues. A blank or non-numeric entry gives no result.

Worked example

Take the matrix with a = 1, b = −2, c = 3, d = 4. Column sums of absolute values: first column |1| + |3| = 4, second column |−2| + |4| = 6. The larger is 6, so the 1-norm is 6.00. Row sums: first row |1| + |−2| = 3, second row |3| + |4| = 7, so the infinity-norm is 7.00. The largest single absolute entry is 4.00. Notice that the two norms differ for the same matrix, because one looks at columns and the other at rows.

Try this example →

Common mistakes

  • Adding the entries with their signs. The minus sign on −2 must be dropped first; summing 1 + 3 or −2 + 4 gives the wrong column total.
  • Mixing up the two norms: the 1-norm uses columns, the infinity-norm uses rows. For a non-symmetric matrix the answers differ.
  • Entering the entries in the wrong positions, for example column by column instead of row by row. This swaps the two norms for a non-symmetric matrix.
  • Expecting the result to be the spectral norm (largest singular value). The 1-norm and infinity-norm are only upper bounds on the largest eigenvalue magnitude, not equal to it.

About this calculator

A matrix norm gives a single number that measures how much a matrix can “stretch” a vector, and different norms answer that question in different ways. For a matrix induced by vector p-norms, the 1-norm equals the largest of the absolute column sums (worst-case stretch when the input is measured with the taxicab / 1-norm), while the infinity-norm equals the largest of the absolute row sums (worst-case stretch under the max-norm) — both quick to compute by hand for a small matrix, unlike the spectral (2-)norm, which needs eigenvalues.

Numerical analysts and engineers use these norms to bound error propagation in linear systems, estimate the condition number of a matrix (how sensitive a solution is to small input errors), and check convergence criteria in iterative solvers. From the four entries of a 2×2 matrix it reads off the 1-norm (largest absolute column sum), the infinity norm (largest absolute row sum) and the largest absolute entry.

Was this helpful?

You can pick up to 3 reactions.

Comments (1)

  • Editorial team

    A quick use is error bounding: the infinity-norm and 1-norm each give an upper limit on how much a 2 by 2 matrix can magnify an input, so the smaller of the two is the tighter bound you can state. For the matrix with entries 1, -2, 3, 4 they are 6.00 and 7.00. Transposing the matrix swaps them, which is a quick way to confirm your entries sit in the right positions. Which norm do you need that is missing here?

Popular calculators

All Calculators