a stochastic process

A single random variable is a one-shot gamble: you flip once and read off a number. But the world rarely holds still — a stock price wanders day after day, a queue grows and shrinks, a molecule jitters through water. A stochastic process is the mathematical object for randomness that unfolds over time: not one random number, but a whole sequence (or continuum) of them, watched as the clock runs.

Precisely, a stochastic process is a collection of random variables { X_t } indexed by a parameter t, all defined on the same underlying probability space. You should picture it two ways at once. Fix the time t and X_t is an ordinary random variable — a snapshot, with its own distribution. Fix instead a single outcome of the underlying randomness (one run of the universe) and the whole family { X_t } becomes one definite function of t, a single trajectory you can plot. The process is the bundle of all these possible trajectories, each weighted by how likely it is.

This double view is the heart of the subject. Questions about a process come in two flavours: marginal questions about one instant (what is the distribution of X_10?) and path questions about the whole journey (will the price ever hit 200? how long until the queue empties?). The grammar in this field — index set, state space, sample path, increments, stationarity, the Markov property — is just the shared vocabulary every particular process (random walk, Poisson process, Brownian motion, Markov chain) is then built from.