When you derive a function from another using the transformation
you're translating the graph of the parent function f(x) horizontally.
More specifically, you translate the graph k units to the left if k is positive, k units to the right if k in negative.
So, starting from the graph of f(x), you have that the graph of
is the same graph of f(x), but shifted 3 units to the left.