1/52
Looks like no tags are added yet.
pivots
the first non-zero entry in a row when a matrix is in row-echelon form (REF)
Any all-zero rows
have to be at the bottom
RREF rules (Reduced row echelon form)
each 1st non zero entry is 1
each pivot is the only non zero # in its column
pivots move right as you go down the matrix
Any all-zero rows have to be at the bottom
Row Echelon Form (REF) rules
rules 2, 3, and 4 of RREF
Rank of a matrix
# of non-all zero rows after RREF, written as rank(A) if matrix A
Coefficient matrix
( coefficients of variables in eqs)
Augmented matrix
( coefficients | constants )
constant and coefficient 2x=5
constant is 5, coeff is 2
Elementary row operations
Swap rows (R1←>R2)
Multiply a row by a constant (R2 → 3*R2)
Add a multiple of one row to another row (R3 → R3+(-2)*R1)
Gaussian Elimination (RREF)
Goal: create leading 1s with zeroes above n below them (you can only use Elementary Row Operations)
Free variable
a variable whose column has no pivot after row reduction
whenever there is a free variable
there is infinite solutions to the system of eqs as long as there is no contradiction (ex: (000|3) is a contradiction)
one solution
independent, one pivot for every variable
no solution
inconsistent, there is a contradiction
inf solutions
dependent, at least one free variable in the system and no contradiction
steps at the end of RREF
check rows for pivots, if you have (0101|2) the 2nd 1 is NOT a pivot
check columns for free variables (there can be more than 1)
m
rows
n
columns
aij
entry in row i, column j
square matrix
matrix w/same # rows and columns
matrices can only be added/subtracted if
theyre the same size
matrix multiplication
row of 1st matrix * column of 2nd matrix → AB=( #1 row * #1 column) + (#2 row * #2 column)
you cn multiply if A=m*r | B=r*n
A*B does not = B*A, dont reverse matrices
Trace
(tr(A) for matrix A), add the #s on the main diagonal, only works for square matrix
Transpose
(A^T for matrix A), switch rows and columns
(A*B)^T = B^T * A^T
Zero matrix
all entries are zero
even tho A+0=A and A-A=0, A*B=0 does NOT mean A=0 or B=0
Identity matrix (I)
1st on the main diagonal and zeroes everywhere else
if the det(A)=0
there is no inverse for matrix A
if the det(A) !=0
there is an Inverse for matrix A
Elementary Matrix is
what you get when you perform ONE row operation on an identity matrix
Invertible Matrix Theorem rules
A is invertible
RREF(A) = I
Ax=0 has only trivial solutions (all variables=0)
Ax=b has one solution for every b
A can be written as a product of Elementary Matrices
Cofactor formula
Cij = (-1)^(i+j) * det(Aij)
Properties of determinants
det(A^T) = det(A)
det(AB) = det(A) *det(B)
det(A^-1) = 1 / det(A)
det(k*A) = k^n * det(A) → n from nxn matrix number
det(In) = 1 → Identity matrix, n is nxn again
det(A^m) = [det(A)]^m
det( k A^m) = k^n * [det(A)]^m
ur doing this stuff to the matrix (ex: multiplying matrix A by k)
properties of inverses
(AB)^-1 = (B^-1) *(A^-1)
(A^T) = (A^-1)^T
(k*A)^-1 = 1/k * A^-1 (k ≠0)
diagonal matrices
every entry outside of the main diagonal is zero
Upper triangular matrix
all entries below the main diagonal are zero
Lower triangular matrix
all entries above the main diagonal are zero
determinant of diagonal matrices
multiply all the numbers on the main diagonal
determinant of triangular matrices
same method as determinant of diagonal matrices
effect of swapping two rows on Determinant of Elementary matrices
changes the sign (±) of the determinant for each swap
effect of multiplying a row by a # on the Determinant of Elementary matrices
det(E) = k ( k is number u multiply by)
effect of Adding a multiple of 1 row to another row on the Determinant of Elementary matrices
determinant doesn’t change
effect of swapping two rows on the determinant of square matrix
changes the sign (±) of the determinant for each swap
effect of multiplying one row by a # on the determinant of square matrix
det(B) = det(A) * k
effect of Adding a multiple of one row to another row on the determinant of square matrix
determinant doesn’t change
u*v < 0
obtuse, theta>90
u*v = 0
right angle, theta = 90
u*v > 0
acute, theta<90
proj a (u)
[ (u*a)/(a*a) ] *a
splitting a vector into perpendicular parts
u = w1 +w2
w1 = proj a (u)
w2 = u - proj a (u)
w1*w2 = 0
For a square system, a nonzero determinant
guarantees exactly one solution
matrix multiplication dimension rule
(m*n)(n*p) = (m*p)
left→ rows
right→ columns
angle eq w vectors
theta = cos^-1 [ (u * v) / ( ||u|| *||v|| ) ]
sin²() + cos²()
1