Loading a Cell Array (.mat file) into Python

Question

I have a 25x4x3 Cell Array which I have saved into a .mat file and now need to load into Python.

A simple .mat file I am familiar with, but being new to Python, this problem is a little over my head. The 25 dimensions and 4 dimensions represent integers that I need to call later in my code. The 3 dimension in the cell array contains a list of datetime arrays, data, and then more character arrays of labels for that data.

e.g. (pseudocode)

.mat(1,1,1) = datetime arrays; 
.mat(1,1,2) = data corresponding to datetime arrays; 
.mat(1,1,3) = labels for each column in that data.

Obviously I need to load this into Python with loadmat('.mat') and then I get confused because my variable explorer does not show the .mat file. I have no idea how to pick apart the contents to correctly convert datettime indices accordingly, numbers to floats, etc.

Any guidance would be appreciated!

Right now I simply have:

filemat = 'Pdata.mat'
Pdata = sio.loadmat(filemat)

It is loading the .mat as an object from what I can tell.

Pdata should be a dictionary, Pdata['mat'] should be the object encoding your mat cell. Because it comes from a cell with mixed data, it won't be a simple numpy numeric array. Previous loadmat questions have demonstrated how to extract information from such structures (see the side bar). You may have to show us the variable (if it's not too big). — hpaulj
– hpaulj, Commented Jul 21, 2017 at 16:35
My last answer along this line: stackoverflow.com/a/42258001/901925; or stackoverflow.com/questions/32723802/… — hpaulj
– hpaulj, Commented Jul 21, 2017 at 16:39

Poelie · Accepted Answer · 2017-07-21 15:38:10Z

2

You're almost there.

First let me build some sample matlab data:

a = num2cell(rand(25,4,3));
a{13} = 'abc';
save('/tmp/tmp4Python.mat','a')

Now load it in Python

import scipy.io
mat = scipy.io.loadmat('/tmp/tmp4Python.mat')

And assign the variable

a = mat['a']

Now if we want to access the 'abc' string we do it as follows (note that Python indexing starts at 0, whereas MATLAB indexing starts at 1):

a[12][0][0]

answered Jul 21, 2017 at 15:38

Poelie

7136 silver badges19 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Tulkkas Over a year ago

how do you deal with it if it is a file save using format v7.3?

user9439906 Over a year ago

@Tulkkas In that case, you can import hdf5storage then hdf5storage.loadmat(file.mat)

Iqbal Jurist · Accepted Answer · 2020-10-20 13:33:15Z

0

How about assigning a variable that only contains a ranged value of 'a'? is it possible like this?

x1 = mat['a', range(1, 20, 1)]

answered Oct 20, 2020 at 13:33

Iqbal Jurist

432 bronze badges

Collectives™ on Stack Overflow

Loading a Cell Array (.mat file) into Python

2 Answers 2

2 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related