# Creating a structure and fields in nested loops

5 views (last 30 days)
Achuan Chen on 12 Dec 2021
Commented: KSSV on 13 Dec 2021
I'm extracting information from a large Exel sheet, it has headers 'category, 'item', 'variety', 'date', 'price', 'unit'; the 'item' header contains fruits and vegetables, and 'variety' contains the different varieties, and 'item' and 'variety' columns are the only 2 concerned in this task.
I'm trying to create a structure with field names 'apples', 'pears', 'carrots, 'cabbage' and for each field it contains an array of distinct varieties of the corresponding item, so it looks something like:
structure=
apples: [distinct varieties for apples]
pears: [distinct varieties for pears] ...
Here's my attempt:
m=1;%m,n,p are counters
n=1;
p=1;
items={'apples','pears','carrots','cabbage'};
matrix=cell(length(items),1);%creates 4-row matrix, where each row
%contains distinct varies of each corresponding item
[A, ia, ic]=unique(data.variety,'stable');
for i=1:length(data.item)
if ismember(data.item{i},items)%checks if it's the item we want
for j=1:length(ia)
if strcmp(data.item{ia(j)},data.item{i})==0 %checks if variety correponds to the correct item
continue
end
matrix{find(strcmp(data.item{i},items)),p}=data.variety{ia(j)};
p=p+1;
end
lst(1).(data.item{i})=matrix{find(strcmp(data.item{i},items)),:};
end
end
My approch was to first obtain a list 'A' of distinct varies for all items, then loop through 'ia' to identity any duplicated varieties for each of the 4 items. If succesful I would obtain a 4-row matrix whose rows are the distinct varieties for the 4 items, then I could assign the corresponding row and structure field.
However, this did not quite work out the way I planned, I think it's giving me a matrix of ALL varities, and the output for the structure is not as anticipated too:
Could someone enlighten me?
Thank you very much!
Achuan Chen on 12 Dec 2021
No not at all, it was the first thing that came to mind and I couldn't figure out another method.

KSSV on 12 Dec 2021
You can use ismember. I don't think you need to use a loop here.
T=readtable('fruitvegprices.csv'); % first column of T is items and second is variety
items={'apples','pears','carrots','cabbage'};
% Get apples
[idx,ia] = ismember(T.(1),items{1});
T_apples = T(idx,:) ; % this gives table of apples alone
Now you can convert it to structure using table2struct.
KSSV on 13 Dec 2021
Yes you need to use unique. Perfect.