Selecting non-zero elements from a symbolic vector

Given a symbolic vector, say [a*b, 0 , a^2*b*c, 1, 0]. How could I obtain a vector from the non-zero entries, I would like for the given example to have [a*b, a^2*b*c, 1] as a result. I tried nonzeros but this does not work for type sym.

[Walter Roberson]

Do you want to remove the entries that are identical to 0, or the entries that can be proven to be 0? For example if one of the entries was 1 - cos(theta)^2 - sin(theta)^2 then we could show that this must be 0 (at least for finite non-nan values), so should it be removed?

Answer 1

[Star Strider]

This works for me (in R2015a):

syms a b c 
V = [a*b, 0 , a^2*b*c, 1, 0];
Vnz = V(V~=0)

Vnz =
[ a*b, a^2*b*c, 1]

[Farid Salazar Wong]

Appreciated, that is all I needed! You are a boss.

[Star Strider]

As always, my pleasure!

Answer 2

[Mischa Kim]

Farid, you could use

syms a b c
d = [a*b, 0 , a^2*b*c, 1, 0];
d(d==0) = []
d = [ a*b, a^2*b*c, 1]

[Farid Salazar Wong]

I get the following message (which I am not sure what it means): Error using symfun>validateArgNames (line 205) Second input must be a scalar or vector of unique symbolic variables.

Error in symfun (line 45) y.vars = validateArgNames(inputs);

Error in sym/subsasgn (line 762) C = symfun(B,[inds{:}]);

[Mischa Kim]

Then you need to post the entire code. Use the paper clip symbol to attach.

Answer 3

[Azzi Abdelmalek]

syms a b c
d = [a*b, 0 , a^2*b*c, 1, 0]
c=nonzeros(d).'

[Farid Salazar Wong]

nonzeros do not work for type sym.

[Star Strider]

The nonzeros function for symbolic arrays was introduced in R2014b.