Concatenate two structures from a starting point

Assume I have two structures with the same fields x and y . Both of them have the fields x.dates and y.dates which are the usual dates in increasing order and x.prices with some prices for each day for a number of variables .In other words if i put 10 number of prices and d1,d2 the number of dates respectively the x.prices has size (d1,10) y.prices has size (d2,10).

I want to concatenate these structures from a starting point in the following way . Assuming the intersection of d1 and d2 is non void and k some place in the intersection :how do I make a structure of the same type which has d dates as the the d1 dates until k and d2 after and similar the prices until k are the prices from x while after they are taken from y ?

Many thanks

[Jan]

A short example would be better then letting the readers guess, what "the usual dates" are.

I do not understand thuis sentence: "if i put 10 number of prices and d1,d2 the number of dates respectively the x.prices has size (d1,10) y.prices has size (d2,10)." What is "a starting point"? The meaning of the 2nd paragraph is not clear to me also.

[Rangesh]

Hi Mihai,

In my understanding, you want to create a new struct having first “k” elements of struct “x” and remaining from struct “y”.

To solve it, first you can find the index “k” at which the intersection of dates of the struct “x” and “y” happens. Then, create a new struct “z” with same fields and store the fields as z.dates=[x.dates(:k) y.dates] and z.prices=[x.prices(:k) y.prices].

You can refer to the following MATLAB Documentations for more information:

Array Indexing - MATLAB & Simulink - MathWorks India

Concatenate arrays - MATLAB cat - MathWorks India

I hope this resolves your query.

Thanks,

Rangesh.

Answer 1

[Jon]

I think this example shows how to do what you are asking for

% Make some example data
d1 = datetime(2023,1,10):days(1):datetime(2023,8,15);
d2 = datetime(2023,7,1):days(1):datetime(2023,10,23);

x.date = d1;
x.prices = rand(numel(d1),10);
y.date = d2;
y.prices = rand(numel(d2),10)

y = struct with fields:
      date: [01-Jul-2023    02-Jul-2023    03-Jul-2023    04-Jul-2023    05-Jul-2023    06-Jul-2023    07-Jul-2023    08-Jul-2023    09-Jul-2023    10-Jul-2023    …    ] (1×115 datetime)
    prices: [115×10 double]

% Find intersection of dates
[C,ia] = intersect(x.date,y.date)

C = 1×46 datetime array
Columns 1 through 14

   01-Jul-2023   02-Jul-2023   03-Jul-2023   04-Jul-2023   05-Jul-2023   06-Jul-2023   07-Jul-2023   08-Jul-2023   09-Jul-2023   10-Jul-2023   11-Jul-2023   12-Jul-2023   13-Jul-2023   14-Jul-2023

Columns 15 through 28

   15-Jul-2023   16-Jul-2023   17-Jul-2023   18-Jul-2023   19-Jul-2023   20-Jul-2023   21-Jul-2023   22-Jul-2023   23-Jul-2023   24-Jul-2023   25-Jul-2023   26-Jul-2023   27-Jul-2023   28-Jul-2023

Columns 29 through 42

   29-Jul-2023   30-Jul-2023   31-Jul-2023   01-Aug-2023   02-Aug-2023   03-Aug-2023   04-Aug-2023   05-Aug-2023   06-Aug-2023   07-Aug-2023   08-Aug-2023   09-Aug-2023   10-Aug-2023   11-Aug-2023

Columns 43 through 46

   12-Aug-2023   13-Aug-2023   14-Aug-2023   15-Aug-2023

ia = 46×1

173 174 175 176 177 178 179 180 181 182

% Set prices before the overlap based on x, and thereafter on y
z.date = [x.date(1:ia(1)-1),y.date];
z.prices = [x.prices(1:ia(1)-1,:);y.prices]

z = struct with fields:
      date: [10-Jan-2023    11-Jan-2023    12-Jan-2023    13-Jan-2023    14-Jan-2023    15-Jan-2023    16-Jan-2023    17-Jan-2023    18-Jan-2023    19-Jan-2023    …    ] (1×287 datetime)
    prices: [287×10 double]