Augmented Matrix: A matrix where each row is an equation, and each column is the coefficients for a particular variable in each equation. For example:
Gaussian Elimination: Using elementary row operations to put a matrix in echelon form
Gauss-Jordan Elimination: Using elementary row operations to put a matrix in reduced echelon form
Reduced Echelon Matrix: A matrix is in reduced row echelon form (also called row canonical form) if it satisfies the following conditions:
- It is in row echelon form.
- The leading entry in each nonzero row is 1 (called a leading one).
- Each column containing a leading 1 has zeros in all its other entries.
Steps: - Make 1
- Make everything below zero
- Repeat with
Free Variable: A variable that can’t be eliminated or constrained. Happens when you have more variables than equations.
LU Decomposition: use elementary row operations to create a lower and upper triangular matrix which together are equivalent to a coefficient matrix
Rank: A way of describing the dimensionality of the output of a transformation. If a transformation outputs a line, the rank is 1, an area: rank 2, and a volume: rank 3. Number of linearly independent columns, or rows, whichever the matrix has fewer of.
Column Space: The range of all possible outputs of a transformation. Equal to the span of the columns of the matrix, if they are treated as vectors.
Column Space Basis:
The columns from the original matrix which have pivots in the REF of that matrix
Row Space Basis:
REF the matrix and pick out the non-zero rows (From the original Matrix (At least in mcmillan))
Nullitity: Columns - Rank
Null Space / Kernel: The set of all inputs to the linear transformation who’s output is the origin.
Null Space Basis: RREF, then write the non zero rows, but with any free variables negative
Cofactor: The determinate of the matrix with the row and column of the mentioned cell removed.
Eigenvector: Becomes a scalar multiple of itself when a given operator is applied to it. Find from eigenvalue using equation
See also, example of complex eigenvector
Eigenvalue: The amount a particular eigenvector get’s stretched
for 2x2 matricies, where m is the mean of the trace: and p is the determinate
See also, example of complex eigenvalue
Diagonalization: Change basis so transformation is a diagonal matrix and becomes light work
- P is change of basis matrix
- D is new diagonal matrix
- For a matrix to be diagonolizable, there need to be at least as many eigenvectors as there are rows
Change of Basis Matrix: To go from any basis to the standard basis, literally put the basis as a matrix. To go from standard to a different basis, write the new basis as a matrix, and calculate it’s inverse. To go between two basis, get the product of the inverse of the second basis X the first basis.
If they aren’t square you do this shit:
Inverses
When determinate is not equal to zero:
If the determinate is zero, there is no inverse.
Elementary Row Operations
Equivalent to operations on systems
Interchange Equations: Swap two rows
Multiply an Equation by a Constant: Multiply one row by a constant.. Ex:
Add a multiple of one equation to another:
Ex: Row 2 + 2* Row 3 Row 2
Application/Problem Types
Wire Temperature
- Make an equation for each node, temp = average of all connected nodes
- solve for constant
- Put into matrix
- RREF
Balance the Chemical Equation
Traffic Management:
- Write an equation for each intersection (A,B,C) where one side is the sum of all the input and the other is the side of all the output
A: - Solve equations for the constant
- RREF
Find the missing values in the partial fraction decomposition
5. Generate a single, distributed equation
6. Separate each “set of coefficients” (here it’s constants and coefficients of x, but it could be powers of x or different variables (I think))
7. RREF
Transformations
one-to-one: if A has linearly independent columns (no free vars)
onto: if A spans all of