ubyte 1.0.1

A package to decode .idx*-ubyte file.

introduction

This package(ubyte 0.1) can decode .idx1-ubyte and .idx3-ubyte files (the MNIST dataset use these forments).
The pakage can use on python3.

functions

This is about how to use the function in this package:

decode_idx3_ubyte(fileDir)

introduction:

decode_idx3_ubyte(fileDir) is a function to decode .idx3-ubyte files and return a array.

argument:

string(fileDir):The path of the file you want to decode.

outprint:

Outprint example:

Getting platform type...
Your computer's platform type is unix-type system .
Start decoding train-images.idx3-ubyte (the complete directory is /xxx/train-images.idx3-ubyte.)
Opening train-images.idx3-ubyte .
 Successfully opened train-images.idx3-ubyte,used time 0.195663 s.
Decoding header information .
 Magic number: 2051,number of images: 60000,size of images: 28*28.
 Successfully decoded header information,used time 0.310604 s.
Decoding main part.
 Decoded 3722 sheets,used time 0.600014 s,etc 9.072753 s.
 Decoded 4370 sheets,used time 0.700014 s,etc 8.911483 s.
 Decoded 5317 sheets,used time 0.800018 s,etc 8.228133 s.
 Decoded 16771 sheets,used time 1.600019 s,etc 4.124406 s.
 Decoded 18187 sheets,used time 1.800016 s,etc 4.138542 s.
 Decoded 36438 sheets,used time 3.000012 s,etc 1.940070 s.
 Decoded 38601 sheets,used time 3.200007 s,etc 1.774134 s.
 Decoded 43789 sheets,used time 3.600013 s,etc 1.332915 s.
 Decoded 46118 sheets,used time 3.800019 s,etc 1.144010 s.
 Decoded 46752 sheets,used time 3.900001 s,etc 1.105301 s.
 Decoded 47665 sheets,used time 4.000001 s,etc 1.035309 s.
 Decoded 49086 sheets,used time 4.200022 s,etc 0.934023 s.
 Decoded 49679 sheets,used time 4.300004 s,etc 0.893515 s.
 Decoded 53669 sheets,used time 4.600003 s,etc 0.542805 s.
 Decoded 55066 sheets,used time 4.800024 s,etc 0.430264 s.
 Decoded 55453 sheets,used time 4.900000 s,etc 0.401964 s.
 Decoded 56246 sheets,used time 5.000012 s,etc 0.333891 s.
 Decoded 57558 sheets,used time 5.200013 s,etc 0.220800 s.
 Successfully decoded main part,used time 5.445718 s.
Successfully decoded train-images.idx3-ubyte,used time 5.544702 s.

output:

numpy.array(images):A numpy array of images.

Output example:

array([[[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]],

       [[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]],

       [[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]],

       ...,

       [[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]],

       [[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]],

       [[0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        ...,
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.],
        [0., 0., 0., ..., 0., 0., 0.]]])

decode_idx1_ubyte(fileDir)

introduction:

decode_idx1_ubyte(fileDir) is a function to decode .idx1-ubyte files.

argument:

string(fileDir):The path of the file you want to decode and return a array.

outprint:

Outprint example:

Getting platform type...
Your computer's platform type is unix-type system .
Start decoding train-labels.idx1-ubyte (the complete directory is /xxx/train-labels.idx1-ubyte.)
Opening train-labels.idx1-ubyte.
 Successfully decoded header information,used time 0.102155 s.
Decoding header information.
 Magic number: 2049,number of images: 60000.
 Successfully decoded header information,used time 0.196112 s.
Decoding main part.
 Decoded 3232 sheets,used time 0.300001 s,etc 5.269510 s.
 Decoded 7921 sheets,used time 0.400006 s,etc 2.630061 s.
 Decoded 8184 sheets,used time 0.500001 s,etc 3.165818 s.
 Decoded 14050 sheets,used time 0.600001 s,etc 1.962366 s.
 Decoded 15233 sheets,used time 0.700002 s,etc 2.057270 s.
 Decoded 21160 sheets,used time 0.800005 s,etc 1.468516 s.
 Decoded 22308 sheets,used time 0.900009 s,etc 1.520752 s.
 Decoded 26843 sheets,used time 1.000010 s,etc 1.235306 s.
 Decoded 26950 sheets,used time 1.100010 s,etc 1.349073 s.
 Decoded 32656 sheets,used time 1.200010 s,etc 1.004883 s.
 Decoded 33839 sheets,used time 1.300005 s,etc 1.005113 s.
 Decoded 38110 sheets,used time 1.400009 s,etc 0.804225 s.
 Decoded 42441 sheets,used time 1.500000 s,etc 0.620662 s.
 Decoded 46652 sheets,used time 1.600003 s,etc 0.457859 s.
 Decoded 55396 sheets,used time 1.800003 s,etc 0.149664 s.
 Decoded 59506 sheets,used time 1.900001 s,etc 0.015837 s.
 Successfully decoded main part,used time 1.956952 s.
Successfully decoded train-labels.idx1-ubyte,used time 2.000741 s.

output:

numpy.array(images):A numpy array of labels.

Output example:

array([5., 0., 4., ..., 5., 6., 8.])

debug

0.2.0

1.The directory different betweens Windows and Unix-like OS .

• Windows:C:\progamefile\
• Unix-like:/Applications

Tips:I'm using macOS 13 .

1.0.0

add requirements.txtto fix the numpy not found error.

others

The package needs numpy in site-package.
Plese visit README.html in the package to get colorful options.