Collect data in structure which contain structure with timestamps

I'm faced to a problem on Matlab, I have a structure DataRobot which contain around 2000 structures with in it 3 datas like this:

DataRobot

------>x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_18_0x2E_840671
----------->Velocity , Position, Angles 
------>x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_015942
----------->Velocity , Position, Angles 
...
..
.  X2000
..
...
------>x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_176554
----------->Velocity , Position, Angles

I would like to collect all the velocity in a simple array but my problem is that the name of the structure

:x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_176554

correspond to a timestamps that I don't really care but I dont wanna write by hand the name of the 2000 timestamps. Is any simple way to collect all my velocity datas?

Thank you very much

[Azzi Abdelmalek]

Make your question clear, how your data are stored? in a text file? post a sample of your data then explain what is your problem

Answer 1

[Stephen23]

Although it is not completely clear, you seem to be describing a scalar structure with multiple fields, each field contains one nested scalar structure with three fields.

First we replicate the nested structures for demonstrating the methods below:

C = {...
      'x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_18_0x2E_840671';...
      'x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_015942';...
      'x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_176554';...
      };
DataRobot = struct();
for k = 1:numel(C)
    tmp = struct('Velocity',sqrt(k),'Position',k,'Angles',k^2);
    DataRobot.(C{k}) = tmp;
end

Method One: structfun

>> structfun(@(s)s.Velocity,DataRobot)
ans =
    1.0000
    1.4142
    1.7321

or if the fields contain non-scalar arrays:

structfun(@(s){s.Velocity},DataRobot)

Method Two: non-scalar structure

Accessing data is trivial when the data has been designed well. In this case there is no real purpose in nesting the structures, and a non-scalar structure would make accessing the data much simpler. So when the structure is flattened into a simple non-scalar structure then your question can be resolved very very simply!

We de-nest these nested structures into a non-scalar array (much easier to work with), and then we can easily extract whatever data you need:

tmp = fieldnames(DataRobot);
S = cell2mat(struct2cell(DataRobot));
[S.Timestamp] = tmp{:};

Now it is easy to access any data:

>> [S.Velocity]
ans =
    1.0000    1.4142    1.7321

You can also access the timetamps in the non-scalar structure:

>> S(2).Timestamp
ans =
x0x32_016_0x2D_Jul_0x2D_13_0x20_17_0x3A_30_0x3A_19_0x2E_015942
>> {S.Timestamp}
ans = 
    [1x62 char]    [1x62 char]    [1x62 char]

Note that accessing the data in a non-scalar structure is significantly faster and simpler than with the nested structures. See the link I gave for more information on non-scalar structures.

Answer 2

[Alex]

Perfect it works well, Thank you very much

[Stephen23]

@Alex: you can also accept my answer, if it solves your question. Accepting answers keeps the volunteers happy :)