We have cumulative distribution function Fx.

Fx = 1/1000x - 1/4

How do we get probability density function (gx) from Fx?

If we have gx, to get Fx, we need to integrate from 0 to x. Am I right?

So to get gx from Fx, we need to find the derivative, because it's an opposite operation from integration.