Time series concepts
A time series as implemented by the Informix® TimeSeries solution contains information about how the data is stored in the table column and information about valid data intervals and where the data is stored on disk.
Understand the following concepts when you create a time series:
- TimeSeries data type
- The data type that defines the structure for the time series data.
- Element
- A set of time series data for one timestamp. For example, a value of 1.01 for the time stamp 2011-1-1 00:45:00.00000 is an element for customer 1001.
- Packed element
- An element in which records for multiple timestamps are stored to save storage space. Packed elements can store hertz data that is recorded at a subsecond frequency and compressed numeric data.
- Timepoint
- The time period for a single element: for example, 15 minutes. In some industries, a timepoint is referred to as an interval.
- Origin
- The earliest time stamp that is allowed. Data that has a timestamp before the origin is not allowed.
- time series instance
- For each TimeSeries data type value, the set of elements that is stored in a container. Each instance has a unique identifier that is stored in the TSInstanceTable table. The time series instance ID is used by some time series routines and SQL statements.The time series instance ID is used by some time series routines.
- Calendar
- A set of valid timepoints in a time series, as specified by the calendar pattern.
- Calendar pattern
- The pattern of valid timepoints and when the pattern starts. The calendar pattern can also specify the length of the timepoint. For example, if you collect electricity usage information every 15 minutes, the calendar pattern specifies that timepoints have a length of 15 minutes, and because you want to collect information continuously, all timepoints are valid.
- Container
- A named portion of a dbspace that contains the time series data for a specific TimeSeries data type and regularity. The data is ordered by time stamp. You can control in which containers your time series data is stored.
- Regularity
- Whether a time series has regularly spaced timepoints or irregularly spaced timepoints.
- Virtual table
- Virtual tables display a view of the time series data in a relational format without duplicating the data. You can use standard SQL statements on virtual tables to select and insert data.