u=x/3. So du=1/3 dx.
So the coefficient of arctan in the solution should be 1/3, not 1/9.
It looks similar to the derivative of inverse tan, so dividing the denominator by 9:
However it doesn't seem to match what Wolfram is telling me. I'm sure I've made a silly mistake, but I'm not seeing it. Any help would be great. Thanks.