2

$$\int \:bx\left(x+a\right)^{n-1}dx$$

I tried using $(x+a-a)$ which gives the apparently correct $$\frac{b}{n+1}\left(x+a\right)^{n+1}-\frac{ba}{n}\left(x+a\right)^n+C$$

However, I tried using by parts as well, using the tabular/DI method. It needed two iterations but my answer is different! My main concern is no $a$ term out the front. By integrating $\left(x+a\right)^{n-1}$ and differentiating $x$, I get $$\frac{bx}{n}\left(x+a\right)^{n+1}-\frac{b}{n\left(n+1\right)}\left(x+a\right)^{n+1}+C$$

Why is this happening?

Ritam_Dasgupta
  • 5,992
  • 2
  • 8
  • 23
user71207
  • 1,543

1 Answers1

6

When you do integration by parts, you get $$b\left[\frac{x(x+a)^{\color{red}{n}}}{n}-\frac{(x+a)^{n+1}}{n(n+1)}\right]+C.$$ (Note: the power in the first term is incorrect in your answer by IP method) We can write this as: $$\frac{b(x+a)^n}{n}\left[x-\frac{(x+a)}{n+1}\right]+C=\color{blue}{\frac{b(x+a)^n}{n(n+1)}\left[nx-a\right]+C}.$$ This is same as the answer you got by the first method (simplify that by factoring out $b(x+a)^n$).

Anurag A
  • 41,067
  • I think I see. To go between the two, is a partial fraction decomposition required? And yes, my integration by parts was a typo – user71207 Jul 26 '21 at 12:29
  • Expanding $\color{blue}{\frac{b(x+a)^n}{n(n+1)}\left[nx-a\right]+C}$ doesnt give you $\frac{b}{n+1}\left(x+a\right)^{n+1}-\color{red}{\frac{ba}{n}\left(x+a\right)^n}+C $ – user71207 Jul 26 '21 at 12:31
  • @user71207 Your first answer can be simplifed as follows: $\frac{b}{n+1}\left(x+a\right)^{n+1}-\frac{ba}{n}\left(x+a\right)^n+C=b(x+a)^n\left[\frac{x+a}{n+1}-\frac{a}{n}\right]+C=\frac{b(x+a)^n}{n(n+1)}\left[nx-a\right]+C$. This last expression is the same as I have in my answer. – Anurag A Jul 26 '21 at 12:34
  • @user71207 If you want to start from $\color{blue}{\frac{b(x+a)^n}{n(n+1)}\left[nx-a\right]+C}=\frac{b(x+a)^n}{n(n+1)}\left[nx+\color{red}{na-na}-a\right]+C=b(x+a)^n\left[\frac{nx+na}{n(n+1)}-\frac{a(n+1)}{n(n+1)}\right]+C$. Now you get the answer you were looking for. – Anurag A Jul 26 '21 at 12:41
  • so you MUST do the $+na -na$ manipulation BEFORE you distribute? – user71207 Jul 26 '21 at 12:43
  • 1
    @user71207 Yes that's done to just match the two forms. – Anurag A Jul 26 '21 at 12:43
  • Interesting. First time I've seen rational functions "hidden" so discreetly from each other. I know it happens in trig a lot when you absorb constants to force an equivalent solution. – user71207 Jul 26 '21 at 13:20