readmat 0.0.1

Helpful functions for loading .mat MATLAB files into Python. Most data types are supported.

readmat

readmat is a Python package with helpful functions for loading .mat MATLAB files into Python. Most data types are supported.

Installation:

pip install readmat

Prerequisites:

python 3
python modules:

• scipy
• numpy

Functions:

load(filename, isNumber=False, isCharArray=False, isStruct=False, isFunction=False, isArray=False, isMatrix=False, isBool=False, isInf=False, isNaN=False, isFunctionHandle=False)

• Example: matdata1 = readmat.load('matfile1.mat', isCharArray=True)
• Usage depends on data type. For structs, an example usage is matdata1.this.is.a.test. For all other data types, matdata1.
• Input: .mat file
• Output: depends on data type

mat2obj(filename)

• Example: matdata2 = readmat.mat2obj('matfile2.mat')
• Example Usage: matdata2.this.is.a.test
• Input: .mat file
• Output: object

mat2dict(filename)

• Example: matdata3 = readmat.mat2dict('matfile3.mat')
• Example Usage: matdata3['this']['is']['a']['test']
• Input: .mat file
• Output: dictionary

options(matobj)

• Example: readmat.options(matdata1)
• Example: readmat.options(matdata2)
• Input: object
• Output: list of options

How structs are loaded:

The load function loads your MATLAB structure as an object in Python. What does this mean? Basically, the method of exploring the depths of your struct in Python is similar to what you'd do in python. Pretty neat examples below:

MATLAB	Python
DWI.dat(1,1,1)	myMat.DWI.dat[0][0][0]
DWI.hdr.private.hdr.test.hello	myMat.DWI.hdr.private.hdr.test.hello
lesion_jhu.mean	myMat.lesion_jhu.mean
data.grades(101)	myMat.data.grades[100]

Known limitations:

The following MATLAB data types are not supported:

• string arrays (for example: "example" [notice the pair of double quotes])
• datetime
• categorical
• table
• timetable

Tips for overcoming each limitation:

(MATLAB commands listed below)

• convert string arrays to char arrays, command: char(stringarray)
• convert datetime to either a char array or a struct, command: char(vardatetime) or struct(vardatetime)
• convert categorical to double, command: double(varcategorical)
• convert table to cell array, command: arrayfun(@convertStringsToChars,arrayfun(@string,[vartable.Properties.VariableNames;table2cell(vartable)]),'UniformOutput',false)
• convert timetable to cell array, command: horzcat(vertcat('Time',cellstr(char(vartimetable.Properties.RowTimes))), arrayfun(@convertStringsToChars,arrayfun(@string,[vartimetable.Properties.VariableNames;table2cell(vartimetable)]),'UniformOutput',false))

Sources cited:

Although I did write part of readmat's code , without the amazing stackoverflow answers from the following stackoverflow users, this project would not have be possible:
users cs01 and jpapon: https://stackoverflow.com/a/29126361
user andyvanee: https://stackoverflow.com/a/6573827

License

This software includes a MIT license.