4CCE1MAP Chap.9 Elimination and Eigenmethods
Elimination and Eigenmethods
Record legal row operations
Gaussian elimination transforms an augmented matrix using row swaps, non-zero row scaling and replacement by a row plus a multiple of another. Upper-triangular form supports back-substitution.
Read pivot evidence
Row-echelon form places leading entries progressively to the right and zero rows at the bottom. Reduced form clears each pivot column above and below.
Pivots, free variables and contradictory rows expose unique, infinite or absent solutions.
Find invariant directions
The eigenproblem Mx=lambda x asks for non-zero directions that a transformation preserves up to scaling.
The characteristic determinant gives candidate eigenvalues, and a homogeneous system gives each eigendirection.
Verify eigenpairs directly
Eigenvectors are defined only up to non-zero scale. Multiplying Mv and comparing with lambda v checks both scalar and vector.
Positive values preserve direction, negative values reverse it, and magnitude describes stretching or contraction.
Make elimination auditable
Choose a non-zero pivot, swap complete rows when necessary and eliminate one column at a time. Copy untouched rows explicitly so that a sign change cannot enter unnoticed. Scaling must apply to the entire row, and the scaling factor must be non-zero.
Recording each operation shows that the transformed equations preserve the original solution set and allows a later reader to locate the first arithmetic drift.
Classify from final rows
A pivot in every variable column supports a unique solution. A missing pivot creates a free variable; if no contradiction appears, parameterise the infinitely many solutions instead of forcing one vector.
An all-zero coefficient row with a non-zero augmented entry asserts an impossibility and proves inconsistency.
Reduced row-echelon form is unique, although different legal routes may reach it, which provides a useful cross-method check.
Build eigenvectors after eigenvalues
The characteristic equation identifies candidate scale factors, but each candidate still needs a non-zero vector in the null space of M minus lambda I.
The zero vector is excluded because it satisfies the scaling relation for every lambda and carries no direction.
Eigenvectors that differ by a non-zero scalar represent the same eigendirection, so normalisation changes presentation rather than the underlying line.
Interpret and verify the transformation
A positive eigenvalue preserves ray direction, a negative value reverses it and the magnitude gives stretch or contraction along the invariant line. Direct multiplication Mv=lambda v checks every component.
For a two-by-two matrix, trace and determinant should equal the sum and product of the eigenvalues, offering compact scalar checks on the characteristic polynomial. These tests complement rather than replace the vector calculation.
What this chapter covers
- 01
Gaussian Elimination
- 02
Row Operations
- 03
Echelon Form
- 04
Pivot Evidence
- 05
Eigenvalues
- 06
Eigenvectors
Worked example · free
Eliminate a three-variable system
- 4Create two lower rows that eliminate x.
- 3Solve the remaining two-variable system.
- 3Back-substitute and verify all equations.
Key terms
- Gaussian Elimination
- Gaussian Elimination creates triangular structure through legal row operations.
- Pivot
- Pivot is a leading entry that identifies a constrained variable column.
- Eigenvalue
- Eigenvalue is the scale factor attached to an invariant direction.
- Eigenvector
- Eigenvector is a non-zero vector retained on its line by a transformation.
Elimination and Eigenmethods FAQ
What makes a row operation legal?
Swap complete rows, multiply a complete row by a non-zero scalar, or replace one row by itself plus a multiple of another. Changing a single entry does not preserve the system.
What does a free variable mean?
It means one variable column lacks a pivot. Without a contradictory row, the system has infinitely many solutions that should be parameterised with an explicit free scalar.
How is an eigenvector checked?
Multiply the matrix by the candidate vector and compare every component with the candidate eigenvalue times that vector. The candidate must be non-zero and dimensionally compatible.
Exam move
Practise elimination with exact fractions and write every row operation. Learn to classify the final rows before back-substitution. For eigenproblems, separate the characteristic equation from the null-space step and always finish with direct multiplication. Use different legal elimination routes on the same system and compare their reduced forms.
Parameterise a free-variable case rather than selecting an arbitrary single vector, and stop immediately when a contradictory row appears. For two-by-two eigenproblems, check the eigenvalue sum against the trace and the product against the determinant before testing each non-zero eigenvector component by component.
Explain the geometric action of each verified eigenvalue by stating whether it preserves direction, reverses it, stretches, contracts or collapses the associated eigendirection.
Working through Elimination and Eigenmethods in 4CCE1MAP? Sia is AskSia’s AI Mathematics tutor — ask any 4CCE1MAP Elimination and Eigenmethods question and get a clear, step-by-step explanation grounded in how 4CCE1MAP is taught and assessed. Read this chapter free, then take your hardest questions to Sia.