So I'm trying to dynamically solve a matrix and have it so I can reuse it next loop however solve seems to return a structure that I can't dynamically cast back into a matrix.
The only code I've found to do it dynamically fails for reason "CELL2MAT does not support cell arrays containing cell arrays or objects."
%Matlab solver test
NumberOfMasses = 20;
lamda = sym('lamda',[NumberOfMasses NumberOfMasses]);
test = lamda * eye(NumberOfMasses,NumberOfMasses);
eqn = lamda * eye(NumberOfMasses,NumberOfMasses) == eye(NumberOfMasses,NumberOfMasses)*2;
ans = solve(eqn,lamda);
cell2mat(struct2cell(ans))%needs to be a matrix for next loop
Is there something I'm missing to Dynamically cast "ans" back to a matrix? (Preferably without writing code that dynamically builds strings and runs getfield() on the struct)

 Réponse acceptée

madhan ravi
madhan ravi le 1 Sep 2019

4 votes

Why not use struct2array() ? Note: Don’t Name a variable named ans .

5 commentaires

Dion Richardson
Dion Richardson le 1 Sep 2019
Modifié(e) : Dion Richardson le 1 Sep 2019
Ahhh thanks, completely missed it. Only issue now is that I have to reshape it every time. Is there anything else I'm missing because this seems like a decent work around. (Just a lot of redundancy in larger code bases and it also fails if the matrix gets turned into a scalar during development)
reshape(struct2array(anz),[NumberOfMasses NumberOfMasses])
I just noticed the warning on ans after I posted hahaha
madhan ravi
madhan ravi le 1 Sep 2019
I don’t see any alternatives at the moment, if you get the desired result using reshape() then why not use it, anyhow you’re willing to use it in a loop??
>> anz = solve(eqn,lamda);
>> anz
anz =
struct with fields:
lamda1_1: [1×1 sym]
lamda2_1: [1×1 sym]
lamda3_1: [1×1 sym]
lamda4_1: [1×1 sym]
lamda5_1: [1×1 sym]
lamda6_1: [1×1 sym]
lamda7_1: [1×1 sym]
lamda8_1: [1×1 sym]
lamda9_1: [1×1 sym]
lamda10_1: [1×1 sym]
lamda11_1: [1×1 sym]
lamda12_1: [1×1 sym]
lamda13_1: [1×1 sym]
lamda14_1: [1×1 sym]
lamda15_1: [1×1 sym]
lamda16_1: [1×1 sym]
lamda17_1: [1×1 sym]
lamda18_1: [1×1 sym]
lamda19_1: [1×1 sym]
lamda20_1: [1×1 sym]
lamda1_2: [1×1 sym]
Look at the order of the fields: all of the _1 appear then all of the _2 and so on. Is that the order that you want the matrix to show up as?
I suggest that instead of converting to array, that you use:
subs(lambda, anz)
Dion Richardson
Dion Richardson le 1 Sep 2019
Modifié(e) : Dion Richardson le 1 Sep 2019
I'm prototyping mathematical functions that I want to later run in C++. My issue is I'm not great at maths so I often make mistakes and change parts of my formula. This occasionally means that in some cases weird conditions come up. The loop time is about 20-30 seconds for some of my code (which is fine for prototyping).
Subs works perfectly! Thanks so much, never would've expected that to be the case.
madhan ravi
madhan ravi le 1 Sep 2019
Thank you sir Walter!

Connectez-vous pour commenter.

Plus de réponses (0)

Catégories

En savoir plus sur Mathematics dans Centre d'aide et File Exchange

Produits

Version

R2019a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by