A time-series is typically a vector
of discrete data samples in which each sample is
comprised of a time-stamp
and a value
is separated by a constant value typically called the
, or period
and each value is rounded
to the nearest value that can be represented
digitally. (See also: sampling rate
There are a wide variety of uses for time-series data, mostly involving digital signal processing
where a signal
is any physical quantity that changes with time.
Because digital signal processing deals with
discrete time-series data, it is possible for the original
signal to have variations which are not described by a
simple discrete value.
Therefore, sometimes time-series data may come
in the form of start,
and end representing the starting value,
the highest value,
and ending value respectively.
An example of this is daily end-of-day stock data which has
an open, high, low, and close value for the whole day.