Langrange interpolation polynomial and Euclidian division

geoffrey159
Messages
535
Reaction score
72

Homework Statement


Let ##x_1,...,x_n## be distinct real numbers, and ## P = \prod_{i=1}^n(X-x_i)##.
If for ##i=1...n ##, ##L_i = \frac{\prod_{j \neq i}^n(X-x_j)}{\prod_{j\neq i}(x_i-x_j)}##, show that for any polynomial A (single variable and real coefs), the rest of the euclidian division of A by P is ##\sum_{i=1}^n A(x_i)L_i ##

Homework Equations



## Q = \sum_{i=1}^n A(x_i)L_i ##

The Attempt at a Solution



Hello, could you tell me if my proof is correct please? It seems correct to me but I have a doubt.

Let ##(B,R)## the quotient and the rest of the euclidian division of A by P. We have that ##\text{deg}(R)<\text{deg}(P) = n ##, and ##A = BP+R##.

Since ## Q(x_i) = A(x_i) = R(x_i)##, then for all ##i = 1...n ##, ##(R-Q)(x_i) = 0 ##, so ##X-x_i## divides ##R-Q##
But for ##i\neq j##, ##X-x_i## and ##X-x_j ## are relatively prime and divide ##R-Q##, so ## P | R-Q##.
However, ##\text{deg}(R-Q) \le \max(\text{deg}(R),\text{deg(Q)}) \le n-1 < n =\text{deg}(P) ##, so ## R-Q = 0 ##
 
Last edited:
Physics news on Phys.org
Seems perfectly fine to me.
You could formulate it in a marginally more direct way perhaps, e.g. saying A-Q is zero at each x_i, hence it divides P, etc. - but this is equivalent to what you say.
 
  • Like
Likes geoffrey159
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...
Back
Top