## 453 Reputation

5 years, 343 days

## Non-commutative algebra is difficult...

@vv The difficulty is that, in eq1, there is both x*y and y*x. To isolate x, one need to have only term like x*y (or equivalently only term like y*x), but not both. So there is no many choice, one need to permute one of those guys. This is a common task performed in non-commutative algebra. The price to pay is that a commutator will appear.

To my knowledge, “solve” is not documented in Physics. Here, I think “solve” should have outputted an error message to not introduce an incorrect result.

## A better solution...

 >
 >
 (1)
 >
 (2)
 >
 (3)
 >
 (4)
 >
 (5)
 >
 (6)
 >

@mskalsi A more elegant solution could be:

CoeffTest := proc (ee, n) options operator, arrow; select(has, ee, g(z)^n)/g(z)^n end proc

## Ambiguity...

@Markiyan Hirnyk There is an ambiguity in the example given by Entvex. X and B could either be vectors or a matrices. Anyway in both cases, the product A*X cannot be commuted. For instance, if X is a vector, X*A has no meaning. So the result is the same, X and B are non-commutative objects in all cases.

## Yes!...

@Markiyan Hirnyk Yes, both X and B are treated as noncommutative operators.

 1 2 3 4 Page 4 of 4
﻿