Let us assume that $Y$ is a discrete r.v. while $Z$ is continuous.
$
X=
\begin{cases} Y & \text{w.p.} && p \\ Z & \text{w.p.} && (1-p) \\
\end{cases}
$
Such a r.v. $X$ has a mixed distribution.
- It is *not discrete*, since the continuous part cannot be mapped with an index.
- It is *not continuous*, since that requires single points to have $0$ probability.
Consequently these mixed distributions do not have a valid [[Probability Mass Function]] or [[Probability Density Function]]. However they always have a [[Cumulative Density Function]].
$ \begin{align}
F_X(x)&=p*\mathbf P(Y \le x)+(1-p)*\mathbf P(Z \le x) \\[2pt]
F_X(x)&=p*F_Y(x)+(1-p)*F_Z(x)
\end{align}
$