Probability density function from Probability distribution function

Working rule. If $F(x)$ is the cdf of a continuous random variable $X$, its pdf is recovered by differentiation:

At the finitely many points where two pieces of $F$ meet and the derivative is not defined (a "corner" of $F$), $f$ may be assigned any convenient value there without affecting any probability — since a single point never carries probability for a continuous $X$ (§11.4.1), the choice at that isolated point is immaterial to every computation. …