Import multiple tables in MATLAB

Question

I need to import many tables (.xlsx format) in MATLAB. I hence tried to do so with a for loop.

The code is reported below:

addpath('C:\...'); % here hide the actual path just for simplicity
numfiles = 50;
mydata = cell(1, numfiles);
for k = 1:numfiles
  myfilename = sprintf('Table_%d.xlsx', k);
  mydata{k} = importdata(myfilename);
end

Even if this actually works it raises problems as duration (hh:mm:ss) and UTCdate arrays are either converted in simple double or NaN.

Can anyone suggest a way around this that solve my problem?

Answer 1

You can likely use the readtable functionality available in the newer versions of Matlab. This creates a matlab table object which should work for what you need.

The option ('DatetimeType','Text') should help you address your time problem (by writing custom code to handle your date format).

Answer 2

In your case I'd rather use the xlsread function. It specifically targets Excel files, hence saving some time compared to readtable or importdata . You can use it in the following way:

Imagine you create a .xls file that looks like this:

First     Second    Third
    1          2        3
    4          5     x    
01:43:00       8        9

Reading those files with xlsread will output three matrices in total.

[num,txt,raw] = xlsread('myExample.xlsx')

num =
     1     2     3
     4     5   NaN
   NaN     8     9

txt = 
   'First'    'Second'    'Third'
   ''         ''          ''     
   ''         ''          'x'    
   '01:43:00' ''          ''    

raw = 
    'First'    'Second'    'Third'
    [    1]    [     2]    [    3]
    [    4]    [     5]    'x'    
   '01:43:00'  [     8]    [    9]

You can customize the import in many different ways, applying filters, ignoring certain data structures etc. I recommend you going through the Matlab documentation for xlsread .