How to solve duration/double Error in yearfrac when forecasting prices?

I'm trying to run a modified Energy Price Simulation shown under https://de.mathworks.com/help/fininst/simulating-electricity-prices-with-mean-reversion-and-jump-diffusion.html#mw_rtc_SimulateElectricityPricesExample_780C89FE

When I want to add back the seasonality as shown in the code snippet to the simulated prices I get the following error message below

I attached the format that is used for the Price dates.

Unable to perform assignment because value of type 'duration' is not convertible to 'double'.

Error in yearfrac (line 96)

YearFraction(Ind) = daysact(Date1(Ind), Date2(Ind)) ./...

Error in Test (line 106)

SimPriceTimes = yearfrac(PriceDates(1), SimPriceDates);

Caused by:

Error using duration/double (line 1075)

Undefined function 'double' for input arguments of type 'duration'. To convert from durations to numeric, use

the SECONDS, MINUTES, HOURS, DAYS, or YEARS functions.

% Add back seasonality.
SimPriceDates = PriceDates(end) + days(0:(nPeriods-1))';
SimPriceTimes = yearfrac(PriceDates(1), SimPriceDates);
CSim = seasonMatrix(SimPriceTimes);
logSimPrices = SimPrices + repmat(CSim*seasonParam,1,nTrials);

[Cris LaPierre]

What do your Price dates represent? Day of year?

What is SimPriceDates supposed to be?

[Tobias Poemsl]

Yes, Price dates represent the day of year. Sim Price Dates is the timespan the simulation is conducted for (one year) and represents the future simulated paths the price forecast should be projected on

Answer 1

[Cris LaPierre]

My suggestion is to turn your PriceDates into datetimes. The yearfrac function can work with those.

Do you know what year the data is from? Here's an example assuming 2019.

PriceDates = readmatrix("Timeseries_Electricity.xls");
PriceDates = PriceDates + datetime(2019,1,0)

PriceDates = 365×1 datetime array
   01-Jan-2019
   02-Jan-2019
   03-Jan-2019
   04-Jan-2019
   05-Jan-2019
   06-Jan-2019
   07-Jan-2019
   08-Jan-2019
   09-Jan-2019
   10-Jan-2019
   11-Jan-2019
   12-Jan-2019
   13-Jan-2019
   14-Jan-2019
   15-Jan-2019
   16-Jan-2019
   17-Jan-2019
   18-Jan-2019
   19-Jan-2019
   20-Jan-2019
   21-Jan-2019
   22-Jan-2019
   23-Jan-2019
   24-Jan-2019
   25-Jan-2019
   26-Jan-2019
   27-Jan-2019
   28-Jan-2019
   29-Jan-2019
   30-Jan-2019

nPeriods = 4;

% Add back seasonality.
SimPriceDates = PriceDates(end) + days(0:(nPeriods-1))'

SimPriceDates = 4×1 datetime array
   31-Dec-2019
   01-Jan-2020
   02-Jan-2020
   03-Jan-2020

SimPriceTimes = yearfrac(PriceDates(1), SimPriceDates)

SimPriceTimes = 4×1

0.9973 1.0000 1.0027 1.0055

[Tobias Poemsl]

Thank you for your help, works like a charm!