How to read and write and store matlab data files in python (*. mat), python. mat

How to read and write and store matlab data files in python (*. mat), python. mat

Background

In the deeplearning process, caffe framework is used, and matlab is generally used to process images (matlab is relatively simple and efficient to process images ), use python to generate the required lmdb file and perform test to generate the result. Therefore, some label information obtained by matlab from image processing will use. the mat file is for python to read, and the result information generated by python also needs matlab for further processing (of course, you can also use txt, so you don't need to handle the structure information yourself ).

Introduction

Data transmission between matlab and python is generally based on the matlab file format. in mat and python, numpy and scipy provide some functions, which can be very good. the data in the mat file is read, written, and processed.

Numpy provides the Array Function to map the Matrix in matlab, while scipy provides two functions: loadmat and savemat to read and write the. mat file.

The following is a simple test program. For more information about function usage, see the help document:

Import scipy. io as SiO2 import matplotlib. pyplot as plt import numpy as np # matlab file name matfn = u'e:/python/Test Program/16225067213162251656_1244.mat 'data = sio. loadmat (matfn) plt. close ('all') xi = data ['xi'] yi = data ['yi'] ui = data ['ui'] vi = data ['vi'] plt. figure (1) plt. quiver (xi [: 5,: 5], yi [: 5,: 5], ui [: 5,: 5], vi [:: 5,: 5]) plt. figure (2) plt. partition F (xi, yi, ui) plt. show () sio. savemat ('saveddata. mat ', {'Xi': xi, 'yi': yi, 'ui': ui, 'vi': vi })

Example 2

Import scipy. io as sioimport numpy as np ### The following describes how to read python. mat file and how to handle the result ### load_fn = 'xxx. mat 'Load _ data = sio. loadmat (load_fn) load_matrix = load_data ['matrix '] # assume that the character variable in the file is matrix. For example, save (load_fn, 'matrix') in matlab '); of course, you can save multiple save (load_fn, 'matrix _ x', 'matrix _ y ',...); load_matrix_row = load_matrix [0] # obtain the first row of matrix in matlab at that time, and arrange the array rows in python ### The following describes how to save python. the mat file is used by the matlab program ### save_fn = 'xxx. mat 'Save _ array = np. array ([1, 2, 3, 4]) sio. savemat (save_fn, {'array': save_array}) # Like above, the first row of the array variable save_array_x = np exists. array ([1, 2, 3, 4]) save_array_y = np. array ([,]) sio. savemat (save_fn, {'array _ x': save_array_x, 'array _ x': save_array_x}) # Similarly, only two different variables are saved

The above is all the content of this article. I hope it will be helpful for your learning and support for helping customers.