Status

All the functions discussed in this document is under a design stage and not yet available.

Overview

Applications

• extract columns of a table for graph plotting
• data value transforms (offset, scale, etc.)
• data reduction (last, mean, …)
• extract a scalar value for HTML substitutions
• select columns of a table or branches of a tree for displaying
• create histogram of a table column
• calculate stats of a histogram
• decode bit flags

Examples

# simple
Temperature->last()->format('%.2f degC')
Temperature->last()->scale(1.8)->offset(32)->format('%.2f degF')
TableData -> last() -> get("NHits") -> histogram(nbins=100,min=0,max=100)

# [TODO] with target
Table->format[column='Temperature']('$.2f degC')

# [TODO] using stack
Temperature->dropBelow(0); Pressure->dropBelow(0); ->zip()
Voltages;Voltages[Main];Voltages[Offset];->sum();->addColumn()

# [TODO] using register file
Temperature->dropBelow(0)->T; Pressure->dropBelow(0)->P; (T,P)->zip()
Voltages[Main]->Vm;Voltages[Offset]->Vo; (Vm,Vo)->sum();(Voltages,.)->addColumn()

Syntax

DATA[ADAPTER1] -> FUNCTOR1(ARG1, ARG2, ...)[ADAPTER2] -> FUNCTOR2(...) ...

• Use a quotation when an identifier includes special character.
• Double quotations are replaced with single quotation on parsing (to be safe in JSON).

Functor Chain, Stack and Register

For functors that takes multiple inputs, stack operators can be used:

CAHNNEL1 -> FUNCTOR1(); CHANNEL2 -> FUNCTOR2() ->...; -> FUNCTOR()

where ; is an operator to push the data into stack.

As an alternative method, output can be pushed back into the data set as a new channel:

CH_1 -> FUNCTOR1()->CH_A; CH_2 -> FUNCTOR2() -> CH_B; CH_A -> FUNCTOR()

Adapters

• Adapters takes a input data and outputs an array of copied values or array of references.
• The scalar functors and array functors need an adapter for their inputs.

• [/]: applies to the entire time-series
• [/PATH] or [PATH]: select a time-series field, table column, tree branch

• [PATH(PATH2=VALUE)]: filter

• @: in-place prefix

Syntax Sugar

• CHANNEL[NAME] will be converted to CHANNEL->get(NAME)

List of Functors

Scalar Functors

• Scalar functors apply to x[k] for each k.
• Unless “in-place” designator is specified, a copy will be created.
• Default adapters are:
  • If x[k] is a histogram, it applies to x[k].counts by default.
  • If x[k] is a graph, it applies to x[k].y by default.
  • If x[k] is a table, a column must be specified using an adapter.
  • If x[k] is a tree, a matching pattern must be specified using an adapter.

• scale(): Scalar<Number> -> Scalar<Number>
• offset(): Scalar<Number> -> Scalar<Number>
• format(fmt): Scalar<Number> -> Scalar<String>
• match(val): Scalar<String> -> Scalar<Bool>
• decode_bits(['aaa','bbb',...]): Scalar<Number> -> Array<String>
• testbit(): Scalar<Number> -> Scalar<Bool>

Array Functors

• Arrays are created by applying an adapter.
• To apply to X[k] for each k as an array, use a target of [X].

• includes(value:String): Array<String> -> Scaler<Bool>
• mean(): Array<Number> -> Scaler<Number> (also: median(), first(), last(), …)
• head(n): Array -> Array (also: tail(n))
• accept_range(min, max) Array<Number> -> Array<Number>
• rescale(percentile=100): Array<Number> -> Array<Number>
• standardize(): Array<Number> -> Array<Number>
• delta(): Array<Number> -> Array<Number>
• sigma(): Array<Number> -> Array<Number>
• histogram(nbins, min, max): Array<Number> -> Histogram

Time-Series Functors

• resample(): TimeSeries -> TimeSeries
• align(): TimeSeries[] -> TimeSeries
• integrate(): TimeSeries -> TimeSeries
• differentiate(): TimeSeries -> TimeSeries

Histogram Functors

• tabulate(columns=['bin_center', 'counts']): Histogram -> Table
• stat(): Histogram -> Tree
• bins(): Histogram -> Tree

Graph Functors

• tabulate(): Graph -> Table

Table Functors

• get(column): Table -> Array
• select(columns, labels=[]): Table -> Table
• range(column, from, to): Table -> Table
• fold(tag_columns): Table -> Tree
• graph(columns, labels=[]): Table -> Graph

Tree Functors

• get(path): Tree -> Scalar
• branch(paths, labels=[]): Tree -> Tree
• rename({old:new, ...}): Tree -> Tree