extract/access same fields from multiple structures (it is a nested structure)

I have 9 structures (1*1 structures), each structure contains 45 fields. some of the field are single values such as : TKE 0.12 in the first structure, TKE 0.005 in the second structure and so on. I am trying to get the TKE fields form these structures and have them in one array (I wanna plot them later). I am not sure how to do it! could anyone please give me a quidance? (by the code: name_of_structure_one.TKE I can get the first TKE which is 0.12 but I want to have all the nine TKE in one array! (TKE is the name of one of the fields)

P.S I put all the 9 structure in one structure using load file_name.mat, thought I can work with it easier maybe! so now I have one structure which has 9 structures in it each structure has 45 fileds in it. how can I get one of the fileds from all 9 structures?

thanks a lot!

Answer 1

[Walter Roberson]

temp = [S1; S2; S3; S4; S5; S6; S7; S8; S9];
all_TKE = [temp.TKE];

The above would only work if the structures all have the same fields in the same order. Otherwise,

all_TKE = cellfun(@(S) S.TKE, {S1; S2; S3; S4; S5; S6; S7; S8; S9});

[Narges Raee]

Thanks Walter so much, all structures have the same fileds in the same order so I have tried the first one, it works perfectly but my structure names are so long (each a brief description of the related experiment) so the first line of the code temp = [s1; s2 ; .....] is so so long, and I need to copy and paste those names 9 times, is it a better way to do that ?

Thanks again man!

[Narges Raee]

I have found a solution that does not nedd the names of the files! I put it here so people who search this question an answer can see it:

c = who ()

this will give you all the structures in a single structure and then you can get the field you are looking for easily!

[Carson Purnell]

For anyone else that searches this question: who() is not a solution. It just makes a cell array of your workspace variable names.

A real solution is to not make 9 different structs, but a single scalar struct of length 9, so you can easily access all the fields of name Z with something like [experiments(:).Z].

Alternatively, depending on the complexity of the data, you could use a cell array and hope that nobody else ever wants to know where things are indexed but again it becomes trivial to access data.

Finally, you might be able to use a table in leiu of a cell array to keep things labeled to a degree.