Intuition The ONE core ideaA system of linear equations is a tangle of variables all mixed together, and the whole topic is one trick: use only reversible moves on the rows to slide the tangle into a staircase shape , so each row reveals one variable at a time. Everything else — pivots, REF, RREF, free variables, inconsistency — is just vocabulary for describing that staircase and reading answers off it.
This page assumes you have seen none of the notation in the parent note . We build every symbol from the ground up, in an order where each idea leans only on the ones before it.
Before matrices, before staircases, there is the linear equation .
Definition Variable and coefficientA variable is an unknown number we want to find — written with a letter like x y z coefficient is a known number multiplying a variable, like the 3 3 x Picture: think of a variable as an empty box □
An equation like
x + 2 y + z = 6 whatever numbers we finally put in the boxes, the left side must total the right side. The word linear means every box appears alone (to the power 1 x 2 x y x
Intuition Why "linear" = "straight"Plot all ( x , y ) x + 2 y = 6 straight line — no bends. That flatness is the entire reason this whole subject stays simple: straight things intersect in predictable ways.
Definition System of linear equationsA system is a set of linear equations that must all hold at the same time. A solution is a choice of numbers for the boxes that keeps every line's promise.
⎩ ⎨ ⎧ x + 2 y + z = 6 2 x + 5 y + 3 z = 15 x + 3 y + 3 z = 11
The curly brace { means "all of these together". Geometrically each equation is a flat sheet in space; a solution is a point sitting on all the sheets at once — an intersection. See Solving Systems of Linear Equations .
Intuition Why we need a new toolThree equations, three boxes, tangled numbers everywhere. Solving by hand-substitution gets messy fast. We want a mechanical, mistake-proof procedure — and for that we first strip away the letters and keep only the numbers. That is the matrix.
A matrix is a rectangular grid of numbers arranged in rows (horizontal) and columns (vertical). We wrap it in big brackets [ ] Picture: a spreadsheet. Row = =
Take the system above. Line up the coefficients by column — all the x y z
1 2 1 2 5 3 1 3 3
Nothing was thrown away except the box-names , and we agreed on their order (x, y, z). The numbers are the equations now.
R i i R i i R 1 R 2 i which row.
The equations also have constants after the = 6 , 15 , 11
Definition Augmented matrixAn augmented matrix glues the coefficient grid and the constants column together, separated by a bar ∣
1 2 1 2 5 3 1 3 3 6 15 11
Common mistake The bar is not "just decoration"Why it feels harmless: it looks like any other column. Fix: the column after the bar holds constants, not a variable's coefficients. A "pivot" landing there later means a contradiction (Section 8) — so the bar changes how you read the final grid.
We now want to simplify the grid without changing which box-values solve it. Only three moves are allowed, and each is chosen because it is reversible (you can undo it), which is exactly why it cannot lose or invent a solution.
c = 0 0 0 = 0 erased a promise, and you can never divide back by 0
Why "add a multiple of one row" is safe: adding equal amounts to both sides of a true equation keeps it true, and you can subtract the same amount to reverse it. Each of these three moves corresponds to left-multiplying by an elementary matrix — but you don't need that yet.
Now the star of the show.
Definition Pivot (leading entry)In a non-zero row, the pivot is the first non-zero number reading left to right. The column it sits in is a pivot column .
Picture: scan a row from the left; the pivot is the first tile that isn't blank (0 0
Intuition Why the pivot mattersA pivot marks the variable that this equation will be "in charge of". If we can make every entry below a pivot zero, that variable vanishes from all lower equations — it's been isolated to the top. Stack these isolations and you get a staircase where each step owns one variable.
A zero row (all 0 no pivot — it makes no promise at all (0 = 0
Definition Row Echelon Form (REF)A matrix is in REF when:
any all-zero rows sit at the bottom ;
each pivot is strictly to the right of the pivot in the row above;
everything below each pivot is 0
Condition 2 is the "down-and-right" staircase; condition 3 is what isolates variables. Reaching REF is exactly Gaussian Elimination .
Intuition What the staircase looks likeDraw a line tracing the pivots: it only ever steps right and down , never back left, never up. Below that line: all zeros. Above and on it: whatever's left. The bottom pivot row is the shortest — often just one variable = a number — the seed you back-substitute upward from.
Definition Reduced Row Echelon Form (RREF)RREF is REF plus :
4. every pivot equals 1 leading 1 );
5. each pivot is the only non-zero entry in its column (zeros above and below).
Reaching this is Gauss-Jordan Elimination . The payoff: each pivot row reads directly as "variable = unique for a given matrix, while REF is not (many staircases, one dusted form).
Once in REF/RREF, count pivots. The number of pivots is the rank .
One solution: a pivot in every variable column, none in the constants column.No solution (inconsistent): a pivot in the constants column — a row saying 0 = nonzero Solving Systems of Linear Equations .Infinitely many: some variable column has no pivot → that variable is a free variable , free to be anything, and everything else follows.
Intuition Why fewer pivots ⇒ freedomEach pivot pins down one variable. If you have more variable-columns than pivots, some columns are un-pinned — those variables float, tracing out a whole line or plane of solutions. Pivot columns, by contrast, are linearly independent directions.
Variables and coefficients
Linear equation = straight line
Elementary row operations
Read off solutions rank freedom
What does the word "linear" force each variable to look like in an equation? It appears alone, to the first power only — no squares, no products, no roots — so the graph is a straight line.
What is a coefficient versus a variable? A coefficient is a known number multiplying a variable; a variable is the unknown box we solve for.
What does R i Row number i
Why is the last column of an augmented matrix drawn after a bar? It holds the constants (right-hand sides), not a variable's coefficients — read differently.
Name the three elementary row operations. Swap two rows; scale a row by a non-zero constant; add a multiple of one row to another.
Why must the scale constant be non-zero? Scaling by 0
Why do all three operations preserve the solution set? Each one is reversible, so no solution can be gained or lost.
What is a pivot? The first non-zero entry (reading left to right) in a non-zero row.
Does a zero row have a pivot? No — an all-zero row makes no promise and contributes no pivot.
What shape do the pivots trace in REF? A staircase stepping only right and down, with zeros below each pivot.
Two extra conditions RREF adds beyond REF? Every pivot equals 1, and each pivot is the only non-zero entry in its column.
Which form is unique — REF or RREF? Only RREF is unique for a given matrix.
What does a pivot in the constants column mean? The system is inconsistent — a row reads 0 = nonzero — so no solution.
What is a free variable? A variable whose column has no pivot; it can take any value.
