Matrix Operation: Why n^2 Steps Needed for Elimination of First Row?

  • Thread starter Thread starter johnchau123
  • Start date Start date
  • Tags Tags
    Matrix
Click For Summary
SUMMARY

The discussion centers on the computational complexity of eliminating the first row in matrix operations, specifically in solving linear equations. It is established that the elimination process requires n^2 operations, where n represents the number of equations and unknowns in an n x n matrix. Participants clarify that each elimination step involves n multiplications, leading to a total of n*n operations for the first row elimination. Alternative interpretations suggest that it may require n*(n-1) operations, but the consensus leans towards n^2 when considering the need for uniformity across rows.

PREREQUISITES
  • Understanding of linear algebra concepts, particularly matrix operations
  • Familiarity with Gaussian elimination techniques
  • Knowledge of computational complexity in algorithm analysis
  • Basic proficiency in mathematical notation and summation formulas
NEXT STEPS
  • Study Gaussian elimination and its computational complexity
  • Learn about matrix multiplication and its implications in algorithm efficiency
  • Explore the derivation of summation formulas for operations in matrix calculations
  • Investigate alternative algorithms for solving linear equations, such as LU decomposition
USEFUL FOR

Mathematicians, computer scientists, and students studying linear algebra or algorithm design will benefit from this discussion, particularly those interested in understanding the efficiency of matrix operations.

johnchau123
Messages
8
Reaction score
0
In my note, it said that

Counting multiplication and division only, in solving linear equations (matrix operation),

Elimination of first row: total n^2 operations

So, forward elimination operations for the matrix is Σ(2 to n) k^2 = n*(n+1)*(2n+1)/6

I have tried to solve the equations but it seem do not need n^2 steps.

Can anyone tell me conceptually why it needs n^2 operations to eliminate the first row?

Thanks.
 
Physics news on Phys.org
You have to eliminate the firsty entry, so you add a multiple of another row - that is n multiplications. Then you need to do the second entry in the row. That is another n multiplications in another row. You do this n times, so that is n*n operations.
 
matt grime said:
You have to eliminate the firsty entry, so you add a multiple of another row - that is n multiplications. Then you need to do the second entry in the row. That is another n multiplications in another row. You do this n times, so that is n*n operations.


But i have the following interpretation

Eliminate the first entry and this is n multiplication
Then, I do it n-1, including the first time.

So, I think it is n*(n-1).

I am quite not sure about this. :confused:
 
To be honest, I'd like you to say what it is that you're doing precisely. I'm not aware of anytime I'd actually want to eliminate the entire first row (of what, by the way? nxn matrix? Why?)
 
matt grime said:
To be honest, I'd like you to say what it is that you're doing precisely. I'm not aware of anytime I'd actually want to eliminate the entire first row (of what, by the way? nxn matrix? Why?)

The following link is a picture which shows what my note says.

http://www.badongo.com/cn/pic/526793"
 
Last edited by a moderator:
Doesn't really answer the questions I asked.

1) you're trying to solve simultaneous equations
2) in how many unknowns and how many equations? I presume n of each.

at least it corrects your first sentence - elimination *for* first row.

Strictly speaking you can do it n*(n-1) operations, I agree. Though you could be supposed to multiply every row by somethings so that they all have the same first entry (eg, 1), and that would be n^2 operations, generically. Unless you describe the algorithm you're attempting to cost, there's no way for anyone else to say what is really going on.
 

Similar threads

  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
13
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 4 ·
Replies
4
Views
1K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 9 ·
Replies
9
Views
3K
  • · Replies 5 ·
Replies
5
Views
12K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 25 ·
Replies
25
Views
3K