Matlab to Python; Converting a hex-address to a decimal numpy array

Question

Its probably done in a few lines but im currently stuck on the following problem:

I want to transfer a matlab operation to python. Matlab looks like this:

startAddress = hex2dec('00D00000');
address = hex2dec(flipud((reshape(dec2hex(address, 8), 2, 4))'));

So i get a 2D-vector with decimal numbers out of this. The closest i got to transfer this to python is the following:

import numpy as np

startAddress = int('00D00000', 16)
address = startAddress.to_bytes(8, 'little')
address = np.frombuffer(addressMsg, dtype=np.uint8)
address = address[0:4].reshape(4,1)

Is this equivalent? The rest of my code doesnt work properly and its quite complex so i wanted to figure out if my mistake is here or somewhere deeper.

Sure! It should be a 4x1 double array (code formatting in comments doesnt break lines; imagine a [0;0;208;0] array, quite simple..) : address = 0 0 208 0 — mnikley
– mnikley, Commented Mar 3, 2020 at 8:49

Rotem · Accepted Answer · 2020-03-02 20:53:12Z

1

MATLAB and Python solutions are equivalent.

The only difference I noticed is that MATLAB result is of class double, and in Python the ndarray elements are of type uint8.
(The type differences are probably not relevant).

There are small issues with the sample code you have posted that makes it non-executable:

The MATLAB code should be: dec2hex(startAddress, 8) instead of dec2hex(address, 8).
The Python code should be: np.frombuffer(address, instead of np.frombuffer(addressMsg,.

You may get the same result using simpler code:

In MATLAB you can use typecast:

address = double(typecast(uint32(startAddress), 'uint8')');

Sample code for testing:

startAddress = hex2dec('12345678');
address = hex2dec(flipud((reshape(dec2hex(startAddress, 8), 2, 4))'));

address2 = double(typecast(uint32(startAddress), 'uint8')');

In Python you can use struct.pack:

address = np.frombuffer(struct.pack("<I", startAddress), np.uint8).reshape(4,1)

Sample code for testing:

startAddress = int('12345678', 16)
address = startAddress.to_bytes(8, 'little')
address = np.frombuffer(address, dtype=np.uint8)
address = address[0:4].reshape(4,1)

# Convert startAddress array of 4 uint8 elements (apply little endian format).
address2 = np.frombuffer(struct.pack("<I", startAddress), np.uint8).reshape(4,1)

MATLAB execution result:

address =

   120
    86
    52
    18

address2 =

   120
    86
    52
    18

Python execution result:

address
array([[120],
       [ 86],
       [ 52],
       [ 18]], dtype=uint8)

address2
array([[120],
       [ 86],
       [ 52],
       [ 18]], dtype=uint8)

answered Mar 2, 2020 at 20:53

Rotem

32.5k6 gold badges40 silver badges81 bronze badges

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

1 Comment

mnikley Over a year ago

Thank you so much ! Youre right with the addressMsg instead of address part, of course, i wrote a function and forgot to mention that address = addressMsg. Trying out your simplification with struct right now :) thank you!

Collectives™ on Stack Overflow

Matlab to Python; Converting a hex-address to a decimal numpy array

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related