Why do I receive this error "Support of character vectors that are not valid variable names or define a number will be removed in a future release. "?

I'm trying to differentiate an expression f which has a horrible coefficient, specifically, (4/3)^(5/6). I want to prevent matlab from giving me the answer to diff(f,x) in the form of an incomprehensible rational number. I can accomplish my objective as follows.

f = @(x) sym('(4/3)^(5/6)')*x^(2/3)

and then

   simplify(diff(f,x))

returns

   (4*2^(2/3)*3^(1/6))/(9*x^(1/3))

which is exactly what I want. However, it's accompanied by a stern warning:

Warning: Support of character vectors that are not valid variable names or define a
  number will be removed in a future release. To create symbolic expressions, first
  create symbolic variables and then use operations on them. 
  > In sym>convertExpression (line 1559)
    In sym>convertChar (line 1464)
    In sym>tomupad (line 1216)
    In sym (line 179)
    In @(x)sym('(4/3)^(5/6)')*x^(2/3)
    In sym>funchandle2ref (line 1308)
    In sym>tomupad (line 1206)
    In sym (line 179)
    In sym/privResolveArgs (line 921)
    In sym/diff (line 21)

I can turn off the warning, but I don't want this functionality to go away, as threatened, in future releases. Is there a way to accomplish the same result without triggering the warning?

Answer 1

[Walter Roberson]

(sym(4)/3)^(sym(5)/6)*x^(sym(2)/3)

[Leo Simon]

Thanks for this @Walter_Roberson. I'm puzzled why only the numerators have to be ``sym-ed'' and not the denominators? Is there a general syntax for these kinds of sym constructions that I can refer to?

[Walter Roberson]

Laziness. A rational value will be constructed as a symbolic rational if either the numerator or denominator or both are symbolic, so you only need to do one of the two. It is permissible to do both.

I tend to instead write something like

Q = @(v) sym(v, 'r');

Q(4/3)^Q(5/6)*x^Q(2/3)

Here, the 4/3 and 5/6 and 2/3 are constructed in double precision, and then the sym() operation approximates the double precision as a rational, reconstructing the fraction. It is a bit more computation work, but a bit less typing for me. But to avoid the potential for the reconstruction to not give back the same fraction, I would probably write

(Q(4)/3)^(Q(5)/6)*x^(Q(2)/3)