make-series operator

Create series of specified aggregated values along a specified axis.

T | make-series sum(amount) default=0, avg(price) default=0 on timestamp from datetime(2016-01-01) to datetime(2016-01-10) step 1d by fruit, supplier

Syntax

T | make-series [MakeSeriesParamters] [Column =] Aggregation [default = DefaultValue] [, ...] on AxisColumn [from start] [to end] step step [by [Column =] GroupExpression [, ...]]

Arguments

  • Column: Optional name for a result column. Defaults to a name derived from the expression.

  • DefaultValue: Default value that will be used instead of absent values. If there is no row with specific values of AxisColumn and GroupExpression, then in the results the corresponding element of the array will be assigned a DefaultValue. If DefaultValue is omitted, then 0 is assumed.

  • Aggregation: A call to an aggregation function such as count() or avg(), with column names as arguments. See the list of aggregation functions. Only aggregation functions that return numeric results can be used with the make-series operator.

  • AxisColumn: A column on which the series will be ordered. It could be considered as timeline, but besides datetime any numeric types are accepted.

  • start: (optional) The low bound value of the AxisColumn for each of the series to be built. start, end, and step are used to build an array of AxisColumn values within a given range and using specified step. All Aggregation values are ordered respectively to this array. This AxisColumn array is also the last output column in the output that has the same name as AxisColumn. If a start value is not specified, the start is the first bin (step) which has data in each series.

  • end: (optional) The high bound (non-inclusive) value of the AxisColumn. The last index of the time series is smaller than this value (and will be start plus integer multiple of step that is smaller than end). If end value is not provided, it will be the upper bound of the last bin (step) which has data per each series.

  • step: The difference between two consecutive elements of the AxisColumn array (that is, the bin size).

  • GroupExpression: An expression over the columns that provides a set of distinct values. Typically it's a column name that already provides a restricted set of values.

  • MakeSeriesParameters: Zero or more (space-separated) parameters in the form of Name = Value that control the behavior. The following parameters are supported:

    Name Values Description
    kind nonempty Produces default result when the input of make-series operator is empty

Returns

The input rows are arranged into groups having the same values of the by expressions and the bin_at(AxisColumn, step, start) expression. Then the specified aggregation functions are computed over each group, producing a row for each group. The result contains the by columns, AxisColumn column and also at least one column for each computed aggregate. (Aggregation that multiple columns or non-numeric results are not supported.)

This intermediate result has as many rows as there are distinct combinations of by and bin_at(AxisColumn, step, start) values.

Finally the rows from the intermediate result arranged into groups having the same values of the by expressions and all aggregated values are arranged into arrays (values of dynamic type). For each aggregation, there is one column containing its array with the same name. The last column in the output of the range function with all AxisColumn values. Its value is repeated for all rows.

Due to the fill missing bins by default value, the resulting pivot table has the same number of bins (that is, aggregated values) for all series

Note

Although you can provide arbitrary expressions for both the aggregation and grouping expressions, it's more efficient to use simple column names.

Alternate Syntax

T | make-series [Column =] Aggregation [default = DefaultValue] [, ...] on AxisColumn in range(start, stop, step) [by [Column =] GroupExpression [, ...]]

The generated series from the alternate syntax differs from the main syntax in two aspects:

  • The stop value is inclusive.
  • Binning the index axis is generated with bin() and not bin_at(), which means that start may not be included in the generated series.

It is recommended to use the main syntax of make-series and not the alternate syntax.

Distribution and Shuffle

make-series supports summarize shufflekey hints using the syntax hint.shufflekey.

List of aggregation functions

Function Description
any() Returns a random non-empty value for the group
avg() Returns an average value across the group
avgif() Returns an average with the predicate of the group
count() Returns a count of the group
countif() Returns a count with the predicate of the group
dcount() Returns an approximate distinct count of the group elements
dcountif() Returns an approximate distinct count with the predicate of the group
max() Returns the maximum value across the group
maxif() Returns the maximum value with the predicate of the group
min() Returns the minimum value across the group
minif() Returns the minimum value with the predicate of the group
stdev() Returns the standard deviation across the group
sum() Returns the sum of the elements within the group
sumif() Returns the sum of the elements with the predicate of the group
variance() Returns the variance across the group

List of series analysis functions

Function Description
series_fir() Applies Finite Impulse Response filter
series_iir() Applies Infinite Impulse Response filter
series_fit_line() Finds a straight line that is the best approximation of the input
series_fit_line_dynamic() Finds a line that is the best approximation of the input, returning dynamic object
series_fit_2lines() Finds two lines that are the best approximation of the input
series_fit_2lines_dynamic() Finds two lines that are the best approximation of the input, returning dynamic object
series_outliers() Scores anomaly points in a series
series_periods_detect() Finds the most significant periods that exist in a time series
series_periods_validate() Checks whether a time series contains periodic patterns of given lengths
series_stats_dynamic() Return multiple columns with the common statistics (min/max/variance/stdev/average)
series_stats() Generates a dynamic value with the common statistics (min/max/variance/stdev/average)

List of series interpolation functions

Function Description
series_fill_backward() Performs backward fill interpolation of missing values in a series
series_fill_const() Replaces missing values in a series with a specified constant value
series_fill_forward() Performs forward fill interpolation of missing values in a series
series_fill_linear() Performs linear interpolation of missing values in a series
  • Note: Interpolation functions by default assume null as a missing value. Therefore specify default=double(null) in make-series if you intend to use interpolation functions for the series.

Example

A table that shows arrays of the numbers and average prices of each fruit from each supplier ordered by the timestamp with specified range. There's a row in the output for each distinct combination of fruit and supplier. The output columns show the fruit, supplier, and arrays of: count, average, and the whole timeline (from 2016-01-01 until 2016-01-10). All arrays are sorted by the respective timestamp and all gaps are filled with default values (0 in this example). All other input columns are ignored.

T | make-series PriceAvg=avg(Price) default=0
on Purchase from datetime(2016-09-10) to datetime(2016-09-13) step 1d by Supplier, Fruit

Makeseries

let data=datatable(timestamp:datetime, metric: real)
[
  datetime(2016-12-31T06:00), 50,
  datetime(2017-01-01), 4,
  datetime(2017-01-02), 3,
  datetime(2017-01-03), 4,
  datetime(2017-01-03T03:00), 6,
  datetime(2017-01-05), 8,
  datetime(2017-01-05T13:40), 13,
  datetime(2017-01-06), 4,
  datetime(2017-01-07), 3,
  datetime(2017-01-08), 8,
  datetime(2017-01-08T21:00), 8,
  datetime(2017-01-09), 2,
  datetime(2017-01-09T12:00), 11,
  datetime(2017-01-10T05:00), 5,
];
let interval = 1d;
let stime = datetime(2017-01-01);
let etime = datetime(2017-01-10);
data
| make-series avg(metric) on timestamp from stime to etime step interval 
avg_metric timestamp
[ 4.0, 3.0, 5.0, 0.0, 10.5, 4.0, 3.0, 8.0, 6.5 ] [ "2017-01-01T00:00:00.0000000Z", "2017-01-02T00:00:00.0000000Z", "2017-01-03T00:00:00.0000000Z", "2017-01-04T00:00:00.0000000Z", "2017-01-05T00:00:00.0000000Z", "2017-01-06T00:00:00.0000000Z", "2017-01-07T00:00:00.0000000Z", "2017-01-08T00:00:00.0000000Z", "2017-01-09T00:00:00.0000000Z" ]

When the input to make-series is empty, the default behavior of make-series produces an empty result as well.

let data=datatable(timestamp:datetime, metric: real)
[
  datetime(2016-12-31T06:00), 50,
  datetime(2017-01-01), 4,
  datetime(2017-01-02), 3,
  datetime(2017-01-03), 4,
  datetime(2017-01-03T03:00), 6,
  datetime(2017-01-05), 8,
  datetime(2017-01-05T13:40), 13,
  datetime(2017-01-06), 4,
  datetime(2017-01-07), 3,
  datetime(2017-01-08), 8,
  datetime(2017-01-08T21:00), 8,
  datetime(2017-01-09), 2,
  datetime(2017-01-09T12:00), 11,
  datetime(2017-01-10T05:00), 5,
];
let interval = 1d;
let stime = datetime(2017-01-01);
let etime = datetime(2017-01-10);
data
| limit 0
| make-series avg(metric) default=1.0 on timestamp from stime to etime step interval 
| count 
Count
0

Using kind=nonempty in make-series will produce a non-empty result of the default values:

let data=datatable(timestamp:datetime, metric: real)
[
  datetime(2016-12-31T06:00), 50,
  datetime(2017-01-01), 4,
  datetime(2017-01-02), 3,
  datetime(2017-01-03), 4,
  datetime(2017-01-03T03:00), 6,
  datetime(2017-01-05), 8,
  datetime(2017-01-05T13:40), 13,
  datetime(2017-01-06), 4,
  datetime(2017-01-07), 3,
  datetime(2017-01-08), 8,
  datetime(2017-01-08T21:00), 8,
  datetime(2017-01-09), 2,
  datetime(2017-01-09T12:00), 11,
  datetime(2017-01-10T05:00), 5,
];
let interval = 1d;
let stime = datetime(2017-01-01);
let etime = datetime(2017-01-10);
data
| limit 0
| make-series kind=nonempty avg(metric) default=1.0 on timestamp from stime to etime step interval 
avg_metric timestamp
[
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0,
1.0
]
[
"2017-01-01T00:00:00.0000000Z",
"2017-01-02T00:00:00.0000000Z",
"2017-01-03T00:00:00.0000000Z",
"2017-01-04T00:00:00.0000000Z",
"2017-01-05T00:00:00.0000000Z",
"2017-01-06T00:00:00.0000000Z",
"2017-01-07T00:00:00.0000000Z",
"2017-01-08T00:00:00.0000000Z",
"2017-01-09T00:00:00.0000000Z"
]