# First principles

1. Jun 2, 2008

### forty

[SOLVED] First principles

Determine the derivatives of the following function from first principles (i.e using the limit
definition of a derivative).

g(x) = x^(3/2)

lim h->0 : (x^(3/2) - (x + h)^(3/2) ) / (-h)

I understand first principles its more so the expansion of the (x + h)^(3/2) that has got me lost.

2. Jun 2, 2008

### Defennder

Use the binomial theorem for non-integer powers. Then expand it for the first few terms, neglect powers of h by truncating the expansion (justified because h approaches zero) then substitute this result into the limit. Something will cancel out and you'll get the answer.

3. Jun 2, 2008

### forty

(x + h)^(3/2)

=> ((x + h)^3))^(1/2)

(x^3 + 2hx^2 + x(h^2) + h(x^2) + 2x(h^2) + h^3)^(1/2)

So i get the following

(x^(3/2) - (x^3 + 2hx^2 + x(h^2) + h(x^2) + 2x(h^2) + h^3)^(1/2)) / (-h)

Firstly is this correct?
Secondly what do i do now? take out a factor form the square root, but what factor? >.<

4. Jun 2, 2008

### Gib Z

The binomial theorem for non-integer powers is a bit of a deeper result than the derivative of a monomial isn't it?

Why not take this approach:

$$\frac{ (x+h)^{\frac{3}{2}} - x^{\frac{3}{2}} }{h} = \frac{ (x+h) \sqrt{x+h} - x\sqrt{x} }{h} = \frac{ x\sqrt{x+h} + h\sqrt{x+h} - x\sqrt{x}}{h}$$

Perhaps you can take it from here =]

5. Jun 2, 2008

### forty

Thank you Gib Z!
As i said i was just having problems with the power 1.5 but after seeing it written like that made everything much clearer.

I multiplied by the conjugate and as a result canceled the h term from the denominator and got the proper result (3/2)x^(1/2).

Thanks!

6. Jun 3, 2008

### Gib Z

Good work, and no problem =] Just in case you are wondering, Defennder's suggestion works as well, though it may be beyond you at the moment.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook