Symbolic dependency apparently sometimes gives wrong answers
5 views (last 30 days)
Mohammad Shojaei Arani on 28 Oct 2022
I have a simple question but could not figour out how to tackle it by myself. Consider the following:
syms p1 y
I am not happy with this since clearly I expected to get the answer [1,0]. I do not know why it behaves like this!!!
Please not that my actual problem are much more complicated and this is just to explain to you my question. So, please do not give me any answer which
just works for this simple example, rather I would like a generic sort of answer.
thanks in advance!
Walter Roberson on 28 Oct 2022
Not a bug. Your first element is a symbolic expression involving the desired variable, but symfundependingon is for unresolved symbolic functions that have the variable as a parameter.
syms f(x) g(x, y) h(y)
expressions involving f or g would be symfun depending on x, but h would not be.
More Answers (2)
John D'Errico on 28 Oct 2022
Edited: John D'Errico on 28 Oct 2022
To be honest, this appears to be more of a bug report than anything else, since that option in hasSymType does not apprear to be working as per the documentation. So I would report it as such, but Answers is not the correct place to report bugs.
Could you find some other solution that would work? Even if someone could find something that works here, it does not get past the fact that hasSymType is the correct tool to solve your specidfic problem, and it should work as per the help.
Paul on 28 Oct 2022
This works, at least for the examples cited in this comment. Haven't tested for anything but those examples.
syms x y z f(x,y)
F = @(expr,var) cellfun(@(s) any(has(s,var)),arrayfun(@(s) findSymType(s,'variable'),expr,'UniformOutput',false));
E = [x y f(x,y)] % then I expect to get the answer [1 0 1].
E =[1 2 x^2;sin(x*y) x z] % then I expect to get the answer [0 0 1;1 1 0].