# How can one change variables in a function?

To be completely honest I am a litte lost.

Let us say a variable x is defined as y = 1/x. I have one function

f[x_]:= x^2


which I would like to multiply into another function

g[y_]:= Exp[y] + y^2 + 3 y


Without having to change all y into x. The way I'd do it is

G[y_]:= f[1/y][Exp[y] + y^2 + 3 y]


Is this a correct approach? Are there smarter ways to do it?

• Use parentheses ((...)) to group expressions, not square brackets ([...]) - those are only for function arguments in Mathematica. Also, why not use f[1/y]*g[y]? Jan 9, 2019 at 17:22
• Yes I can do that. I just wanted to make sure that just putting f[1/y] inside there wouldn't cause any issues. Thank you ! Jan 9, 2019 at 20:03

There are quite a few ways to accomplish what you want to do. Here are some.

f[x_] := x^2
g[y_] := Exp[y] + y^2 + 3 y


The simplest is probably is what Lukes suggested.

fg1[y_] := f[1/y] g[y]


or as pure function

fg2 = f[1/#] g[#] &;


However, in this case, since it is easy to define x as the inverse function of y,

x[y_] := 1/y


Composition ( @* ) can also be used.

fg3[y_] := (f@*x)[y] g[y]


or

fg4 = f@*x@# g@#&;


The last definition is a good example of the Wolfram Language's terse functional style. Some consider it elegant; others consider it obscure.

You can be satisfy yourself that all the definitions represent the same function by evaluating

Show[

• You seem to have forgotten a space in before g in the definition of fg4 Jan 9, 2019 at 20:18