Statement
To find the gradient $ m $ of any function $ f $ at point $ x $, you use:
$$ m = \lim_{h \to 0}\left(\frac{f(x + h) - f(x)}{h}\right) $$
The average increase between two points
To get the average increase between points $ A $ and $ B $, you use:
$$ m = \frac{\Delta y}{\Delta x} = \frac{y_A - y_B}{x_A - x_B} $$
When $ y = f(x) $, you get the following.
$$ m = \frac{f(x_A) - f(x_B)}{x_A - x_B} $$
The gradient at one point
To get the gradient/slope at one point, we must choose another point that is very close to the first point.
When $ x_A $ is the point of which we want to know the gradient/slope, the other point can be $ x_A + 0.1 $ (the smaller the better). Note that the difference between $ -0.01 $ and $ 0.01 $ in the denominator is negligible.
$$ m = \frac{f(x_A) - f(x_A + 0.1)}{x_A - (x_A + 0.01)} = \frac{f(x_A) - f(x_A + 0.1)}{0.01} $$
The closer the two are to each other, the more accurate the calculation becomes. Yet, we can't use $ 0 $, since dividing by $ 0 $ is undefined.
Using limits
Although we can't use $ 0 $, we can approach $ 0 $ using limits.
A limit is a way to say that a variable approaches a value while never actually being that value.
$$ m = \lim_{h \to 0}\left(\frac{f(x + h) - f(x)}{h}\right) $$
- This is the notation for limits. Write $ \lim $ and beneath it the variable and the value it almost but not quite reaches.
- When working with derivatives, we often use $ h $ for the almost-zero variable.
- Note that $ h $ can't be zero since you'll be dividing by zero then, and that is undefined.
Notation
The derivative has its own notation. The gradient/slope of $ f(x) $ is often written as $ f'(x) $. The definition then becomes:
$$ f'(x) = \lim_{h \to 0}\left(\frac{f(x + h) - f(x)}{h}\right) $$
This function $ f' $ is called "the derivative of $ f $", or just "the derivative".