Functions & Operators

The below tables list the functions and operators available for Driver Models.

Operators

Type	Operator	Notes
Arithmetic	+ - * /	The standard arithmetic operators
Arithmetic	%	modulus, or remainder
Arithmetic	^ **	“to the power of”. x^y can be written as x**y or pow(x,y)
Logical Boolean	&& \|\| !	And, Or, Not
Bitwise	& \| xor	And, Or, Xor
Comparison	> < >= <=	Greater than, Less Than, Greater Than or Equal To, Less Than or Equal To
Comparison	= ==	Is Equal To
Comparison	!=	Is Not Equal To
Constants	e pi	Euler’s constant, approximately 2.7182818, Pi, approximately 3.14159265

Functions

Math

Function	Notes
`abs(Number)`	Returns the absolute value of the specified number
`acos(Number)`	Returns the angle whose cosine is the specified number
`asin(Number)`	Returns the angle whose sine is the specified number
`atan(Number)`	Returns the angle whose tangent is the specified number
`ceiling(Number)`	Returns the smallest integral value that is greater than or equal to the specified number
`cos(Angle)`	Returns the cosine of the given angle
`cosh(Angle)`	Returns the hyperbolic cosine of the specified angle
`exp(Number)`	Returns e raised to the specified power
`floor(Number)`	Returns the largest integer less than or equal to the specified number
`ln(Number)`	Returns the natural (base e) logarithm of a number
`log10(Number)`	Returns the natural (base 10) logarithm of a number
`log(Number, Base)`	Returns the logarithm of a number in a specified base
`pow(Number, Power)`	Returns a number raised to the specified power
`rand()`	Returns a random number between 0 and 1
`rangepart(value, min[, max])`	Returns that part of a value that lies between min and max)
`round(Number[, d])`	Rounds the argument, (optionally to the nearest ’d’ decimal places)
`sin(Angle)`	Returns the sine of the given angle
`sinh(Angle)`	Returns the hyperbolic sine of the specified angle
`sqrt(Number)`	Returns the square root of the specified number
`tan(Angle)`	Returns the tangent of the given angle
`tanh(Angle)`	Returns the hyperbolic tangent of the specified angle
`trunc(Number)`	Rounds the specified number to the nearest integer towards zero

Aggregate

Function	Notes
`avg(A, B, C, ..n)`	Returns the average of the specified numbers
`max(A, B, C, ..n)`	Returns the maximum of the specified numbers
`min(A, B, C, ..n)`	Returns the minimum of the specified numbers
`sum(A, B, C, ..n)`	Returns the sum of the specified numbers

Time-period Functions

Function	Notes
`asset(AssetName, AttributeName)`	Returns the value of an attribute on an asset
`simulationruntime()`	Returns the simulation time (from the active scenario)
`periodCurrent()`	Gets the current 0 based period number
`periodAverage(N[,[start][,end][,error_value])`	Returns the cumulative average of the argument over all the model iterations to date, or between start and end if set. Error value applies prior to values prior to the range, or if not set the period 0 value
`periodWeightedAverage(N, W[,[start][,end][,error_value])`	Returns the cumulative weighed average of the argument over all the model iterations to date, or between start and end if set. Error value applies prior to values prior to the range, or if not set the period 0 value
`periodCount([N,[start][,end])`	Returns the number of model iterations to date, or between start and end if set.
`periodFirst(N)`	Returns the value of the argument at the first model iteration.
`periodLast(N)`	Returns the value of the argument at the last model iteration.
`periodMax(N[,start][,end][,error_value])`	Returns the high-water-mark of the argument over all the model iterations to date, or between start and end if set. Error value applies prior to values prior to the range, or if not set the period 0 value
`periodMin(N[,start][,end][,error_value])`	Returns the low-water-mark of the argument over all the model iterations to date, or between start and end if set. Error value applies prior to values prior to the range, or if not set the period 0 value
`periodSum(N[,start][,end][,error_value])`	Returns the cumulative sum of the argument over all the model iterations to date, or between start and end if set. Error value applies prior to values prior to the range, or if not set the period 0 value
`periodOpeningBalance(I[,A[,R]])`	Returns the start-of-period balance, based on an Initial balance and accruing Additions and Removals on each subsequent period.
`periodClosingBalance(I[,A[,R]])`	Returns the end-of-period balance, based on an Initial balance and accruing Additions and Removals on each period.
`periodPresentValue(d, N)`	Returns the present value of parameter N using the specified discount rate of d
`periodNPV(d, N)`	Returns the NPV of parameter N using the specified discount rate of d
`periodDelay(p, N[, default])`	Returns the value of the argument N, but p periods later. Prior to that, it returns optional [default], else 0
`periodVar(N)`	Returns the variance of the argument N
`relativePeriodSum([node], N)`	Returns the period sum of “node” over a rolling window into the past. The value of “N” determines the number of periods to go into the past including the current period.

Time-period Functions for All Periods

Function	Notes
`periodAverageAll(N)`	Returns the average of the argument over all of the time periods
`periodCountAll()`	Returns the total number of periods
`periodLastAll(N)`	Returns the value of the argument in the last period
`periodMaxAll(N)`	Returns the maximum value of the argument over all of the time periods
`periodMinAll(N)`	Returns the minimum value of the argument over all of the time periods
`periodSumAll(N)`	Returns the sum value of the argument over all of the time periods
`periodWeightedAverageAll(N, W)`	Returns the weighted average of the argument N over all of the time periods, weighted by W

Conditional Functions

Function	Notes
`if(test,truepart,falsepart)`	If ’test’ is true, returns ’truepart’, else ‘falsepart’
`iferror(calculation, errorresult)`	If ‘calculation’ has an error, return the error result, otherwise return the result of ‘calculation’
`switch(value, condition, result, condition2, result2, condition3, result3, ...)`	Evaluates a value, then returns the result associated with the value

Conditional Time-period Functions

Function	Notes
`AverageIf(condition, N)`	Returns the cumulative average of the argument over all the model iterations to date if the condition is met
`CountIf(condition)`	Returns the number of model iterations to date if the condition is met
`SumIf(condition, N)`	Returns the cumulative sum of the argument over all the model iterations to date if the condition is met

Execute Functions

Function	Notes
`execute(model name, period, input1Name, [input1Node], input2Name, [input2Node], ...)`	Executes an Akumen app (use period for the period number in driver models, 0 for Py/R). Returns 0 if successfull, otherwise -1. Note that this is not intended for large Py/R models, and could cause a performance impact. It is designed to be used as simple helper functions the driver model cannot perform. It is also limited to int/float inputs only
`executeoutput([executeNode], output1Name)`	Returns the result of the execute function, getting the value output1Name from the result. This is limited to simple int/float outputs only

Optimise Functions

Function	Notes
`optimise(numParticles, maxIterations, [evaluationnode], targetvalue, [first_input], first_input_lower_bound, first_input_upper_bound, ... to max of 5 inputs)`	Optimises a node, altering inputs until the target value is reached
`optimiseerror([optimise node])`	Returns the error, ie how far from the target value was reached
`optimiseInput([optimise node], [Input node])`	Returns the final input value from the optimisation
`minimise(numParticles, maxIterations, [evaluationnode], [first_input], first_input_lower_bound, first_input_upper_bound, ... to max of 5 inputs)`	Alters inputs until the minimal value of the result node is reached
`minimiseInput([minimise node], [Input node])`	Returns the final input value from the minimisation
`maximise(numParticles, maxIterations, [evaluationnode], [first_input], first_input_lower_bound, first_input_upper_bound, ... to max of 5 inputs)`	Alters inputs until the minimal value of the result node is reached
`maximiseInput([maximise node], [Input node])`	Returns the final input value from the maximisation

Datasource Functions

See here for information on setting up datasources, or here for further information on how to use datasources within Value Driver Models.

Function	Notes
`datasource([datasource_node], value_column, aggregation_method, forwardfill, forwardfillstartingvalue, filter1column, filter1value, filter2column, filter2value, ...)`	Links a calculation to the datasource and applies an optional filter
`datasourceall([datasource_node], value_column, aggregation_method, forwardfill, forwardfillstartingvalue, filter1column, filter1value, filter2column, filter2value, ...)`	Links a calculation to the datasource and applies an optional filter

value_column: The column within the datasource that contains the value
aggregation_method: sum, avg, min, max, count, stddev, firstquartile, thirdquartile, firstknownvalue, lastknownvalue or median
forwardfill: true/false/number: If true, the last known value is used to forward fill when there is no data. False will cause an error if there is no data. Number will use the number value where there is no data
forwardfillstartingvalue: A number to use if there is no last know value to start with
filter1column: The column that contains the string value to filter the data on for this particular calculations
filter1value: The string value to filter on

Info

Datasource specific filtering can be used across multiple columns in the datasource by adding in filter2column, filter2value, filter3column, filter3value etc.

Info

datasourceall can also be used in place of datasource (using the same parameters). Instead of operating at a single time period, it operates across the entire dataset (honouring the load id and additional filter). This allows you to do things like get the stddev or mean of the entire dataset.

Timeseries Functions

If only one row exists in the timeseries node, data can be referenced in the calculation formula by only providing the timeseries node name.
If multiple rows exist, the row name will need to be specified in the calculation formula as well. See grid below for calculation formula formats.

Function	Notes
`timeseries([timeseries_node])`	References time series node with name “timeseries_node”
`timeseries(timeseries([timeseries_node],"row_name"))`	References time series node with name “timeseries_node” and row label “row_name”
`timeseries([timeseries_node],[numeric_node_name])`	References time series node with name “timeseries_node” and the row set to the 0-based value “numeric_node_name”

DateTime

Function	Notes
`CurrentDayInMonth()`	The numerical day of the month of the current period
`CurrentDayInWeek()`	The numerical day of the week of the current period
`CurrentDayInYear()`	The numerical day of the year of the current period
`CurrentMonth()`	The numerical current month
`CurrentPeriodDateTimeUtc()`	The current period datetime in Excel format (ie 40000)
`CurrentYear()`	The numerical current year
`DaysInMonth()`	The numerical days in the month
`FirstPeriodDateTimeUtc()`	Period 0’s datetime in Excel format
`LastPeriodDateTimeUtc()`	The last period’s datetime in Excel format
`NumPeriods()`	The number of periods in the model

Random

Function
`ContinuousUniform(lower, upper, [seed])`
`Lognormal(mu, signma, [seed])`
`Normal(mean, stddev, [seed])`
`Pert(min, mostlikely, max, [seed])`
`Triangular(lower, upper, mode, [seed])`

Info

The seed is optional, and applies at time period 0. This means that the random number generated for time period 0 will always be the same as long as the seed remains the same. The following periods reuse the random number generator meaning the pattern of numbers will be exactly the same for each execution. This guarantees consistency of results. The seed can also be applied using a scenario parameter.

Range Checking

Click here for more information on range checking nodes.

Function	Notes
`rangecheck(sourcenode, lowlow, low, high, highhigh)`	Performs a check against of the source nodes value against the limits
`rangecheckresult(rangechecknode)`	Returns a value corresponding to which limit has been broken. -2 = lowlow, -1 = low, 0 = none, 1 = high, 2 = highhigh