apply (Series)

Apply function to time series period by period

Syntax

x = apply(func, x, dates, y1, y2, etc...)

Input arguments

func [ function_handle ]

Function that will be applied to the input series x period by period (allowing for time recursive formulas); see Description for details of referencing the input time series (x, y1, y2, etc...) in the func.

x [ Series ]

Input series to which the func will be applied period by period.

dates [ Dater ]

Dates at which the new values will be calculated for x and stored in the output series.

y1, y2, etc... [ Series | numeric ]

Extra input arguments with which the func will be evaluated.

Description

The function func must accept a total of N input arguments where N is 2 plus the number of the extra input arguments y1, y2, etc... The first two input arguments must be the input series x and the time index, say t.

Any use of a time series in the func must be followed explicitly by a round-bracket reference (t, :) where t is the name of the time reference. In a special case where the time series is a scalar series (a single column), this can be abbreviated to (t).

Example

Autoregressive process

>> x = Series(qq(2020,1), 1);
>> func = @(x, t) 0,8*x(t-1) + 5;
>> x = apply(func, x, qq(2020,2):qq(2021,4))
x = 
    Series Object: 8-by-1
    Class of Data: double
    2020Q1:        1
    2020Q2:      5.8
    2020Q3:     9.64
    2020Q4:   12.712
    2021Q1:  15.1696
    2021Q2:  17.1357
    2021Q3:  18.7085
    2021Q4:  19.9668
    'Dates'    ''
    User Data: Empty

Example

Autoregressive process with exogenous input

>> x = Series(qq(2020,1), 1);                
>> y = Series(qq(2020,1):qq(2021,4), (1:8)');
>> func = @(x, t, y) 0.8*x(t-1) + y(t);      
>> x = apply(func, x, qq(2020,2):qq(2021,4), y)
x = 
    Series Object: 8-by-1
    Class of Data: double
    2020Q1:        1
    2020Q2:      2.8
    2020Q3:     5.24
    2020Q4:    8.192
    2021Q1:  11.5536
    2021Q2:  15.2429
    2021Q3:  19.1943
    2021Q4:  23.3554
    'Dates'    ''
    User Data: Empty