Time series data types

Time series functions operate on a wide variety of special data types designed especially for time series data.

Table 1. Time series data types. These are binary data types especially designed to hold time series data. Alternatively, you can store time series data in any BinaryType data array, such as those offered by Parquet, Avro, ORC, or JSON. CSV does not offer a suitable BinaryType data array for storing time series data.
Data type Description
DoubleTimeSeries Time series with univariate observations of type Double, for example:
[(1, 7.2), (3, 4.5), (5, 4.5), (5, 4.6), (5, 7.1), (7, 3.9), (9, 1.1)]
DoubleArrayTimeSeries Time series with multivariate observations of type Double, for example:
[(1, [7.2, 8.74]), (3, [4.5, 9.44]), (5, [4.5, 10.12]), (5, [4.6, 12.91]), (5, [7.1, 9.90]), (7, [3.9, 3.76])]
DoubleSegmentTimeSeries DoubleTimeSeries that has been segmented, for example:
[(1,[(1, 7.2), (3, 4.5)]), (5,[(5, 4.5), (5, 4.6), (5, 7.1)]), (7,[(7, 3.9), (9, 1.1)])]
DoubleArraySegmentTimeSeries DoubleArrayTimeSeries that has been segmented, for example:
[(1,[(1, 7.2, 8.74), (3, 4.5, 9.44)]), (5,[(5, 4.5, 10.12), (5, 4.6, 12.91), (5, 7.1, 9.90)]), (7,[(7, 3.9, 3.76)])]
StringTimeSeries Time series with univariate observations of type String, for example: 
[(1, "a"), (3, "b"), (5, "c"), (5, "d"), (5, "e"), (7, "f"), (9, "g")]
StringArrayTimeSeries Time series with multivariate observations of type String, for example: 
[(1, ["a", "xq"]), (3, ["b", "zr"]), (5, ["c", "ms"]), (5, ["d", "rt"]), (5, ["e", "wu"]), (7, ["f", "vv"]), (9, ["g", "zw"])]
StringSegmentTimeSeries StringTimeSeries that has been segmented, for example:
[(1,[(1, "a"), (3, "b")]), (5,[(5, "c"), (5, "d"), (5, "e")]), (7,[(7, "f"), (9, "g")])]
StringArraySegmentTimeSeries StringArrayTimeSeries that has been segmented, for example:
[(1,[(1, ["a", "xq"]), (3, ["b", "zr"])]), (5,[(5, ["c", "ms"]), (5, ["d", "rt"]), (5, ["e", "wu"])]), (7, [(7, ["f", "vv"]), (9, ["g", "zw"])])]
Table 2. Pseudo data types. These are not actual data types, but are placeholders used by the documentation to represent any of several time series data types.
Placeholder Description
TimeSeries Any of the following data types:
  • DoubleTimeSeries
  • DoubleArrayTimeSeries
  • DoubleSegmentTimeSeries
  • DoubleArraySegmentTimeSeries
  • StringTimeSeries
  • StringArrayTimeSeries
  • StringSegmentTimeSeries
  • StringArraySegmentTimeSeries
SegmentTimeSeries Any of the following data types:
  • DoubleSegmentTimeSeries
  • DoubleArraySegmentTimeSeries
  • StringSegmentTimeSeries
  • StringArraySegmentTimeSeries
ArrayTimeSeries Any of the following data types:
  • DoubleArrayTimeSeries
  • DoubleArraySegmentTimeSeries
  • StringArrayTimeSeries
  • StringArraySegmentTimeSeries
Table 3. Time reference system data type
Data type Description
TRS A time reference system (TRS) defines the granularity of time series timeticks, and defines a start time that maps a particular time series timetick to a corresponding moment in time.