Re: how to solve this error mesages from this program?

[Colin Macdonald]

On 2019-02-07 7:23 a.m., Doug Stewart wrote:
>   There seems to be a problem with the eval() function when it is used with
> the results from a symbolic calculation.
>   The function given above calculates a symbolic matrix which it is
> then trying to convert it to a numeric answer. I tried to use double 
> instead of
> eval but it still did not work. I then tried subs() to convert each 
> symbolic symbol
> to a number and that also failed. In the end I had to convert it to a 
> function_handle
> and then evaluate it at the given values.

Thanks!   I have filed:

https://github.com/cbm755/octsympy/issues/933

It would be great if someone can test the attached script in a somewhat 
recent version of Matlab (with the Symbolic Toolbox) and report the 
results, either via email or directly to the bug report.

> Should the eval function work for this case?

I'm not sure.  I can't find documentation for @sym/eval from the 
Mathworks, perhaps its somewhat deprecated?  My theory is it should do 
the same as `double(subs(f))`.

`subs(f)` is known not to work (and the error message should make this 
clear).  But if people are using these things, then `subs(f)` seems 
semantically better than `eval(f)` so I've filed

https://github.com/cbm755/octsympy/issues/934

These sorts of non-explicit commands are pretty magical (dig around in 
your workspace, find value of variables; I'm not a fan).

Colin

symevaltest.m
Description: Text Data