How to write a changing table name in setdiff function?

Hello,

I would like to know how to write a function inside which there is a changing table name.

I have created a table which will change at each loop

eval(['Port' num2str(i-1) '= Portfolio'])

At some point I need to make a difference between this table and a new portfolio so i used the function setdiff

for example, there is a Port1 which is created, now I need to execute this :

[i,j] = setdiff(Port1(:,2), Portfolio(:,2))

Port1 is created by the eval function. Can you tell me how I can incorporate the name of the table dynamically inside the setdiff function which will pick up automatically the newly created table ( port2, port3....). I tried this but its not working :

   [i,j] = setdiff(['Port' num2str(i-1)](:,2),Portfolio(:,2))

Thank you very much for your help

D

[Stephen23]

@Davin: it is almost always a bad idea to create variable dynamically. Read this to know why:

https://www.mathworks.com/matlabcentral/answers/196952-how-to-create-new-variables-in-batches-with-strcat

Much simpler and more reliable is to use indexing. You can easily put the tables into cell arrays, and using indexing is then trivial.

Answer 1

[Guillaume]

Despite what's been suggested, there is no need to convert the tables to cell arrays (and lose information such as column and row names). Just put the tables into a cell array.

Numbered variables are a clear indication that what is in the variables should be stored together in one container, not as individual variables. In this case a cell array of tables would be the most appropriate.

portnumber = 5;
Port{portnumber} = Portfolio; %store in cell array
[coldiff, loc] = setdiff(Port{portnumber}(:, 2), Portfolio(:, 2))

No eval needed, just straightforward indexing and you still get to manipulate tables with all the advantages that they bring.

[Davin]

Its exactly what I did, and its far much simpler.

Thanks Guillaume.