tidserie
Tidserie, or time series, is a sequence of data points indexed in time, often measured at regular intervals such as seconds, minutes, days, or years. Each observation y_t represents the value of a variable at time t. Time series can also be irregular, but most common methods assume a consistent sampling frequency. The data are typically ordered so that temporal dependence may exist, meaning observations closer in time are often more related than distant ones.
A time series usually comprises several components: a long-term trend reflecting persistent increase or decrease, seasonal
Modeling and forecasting rely on distinguishing stationary from non-stationary series. Stationarity implies constant mean and variance
Practical considerations include handling missing data, irregular sampling, outliers, and structural breaks. Evaluation typically uses time-ordered