C-like data types in µPy

[AllanGErnst]

Hello everybody

I’m a newbie with Python and MicroPython, so please be gentle
I’m making an embedded system using a stm32f429i-disc1 board running MicroPython.

I would like to make a 2D array but I found that the overhead is massive, if I use native MicroPython, so I’ve had a look at the uctypes library. Unfortunately I can’t really get my head around it.

Could somebody please tell me how to get as close to the c type “int twodim[5][10];” as possible? With a minimum amount overhead both with respect to time and memory.

Please include the top of the code, with the imports.

Thanks a lot

Best regards Allan

[OutoftheBOTS_]

Will this possibility give you what you want http://docs.micropython.org/en/latest/w ... array.html

[pythoncoder]

What sort of data do you want to store in the array? 32 bit integers? 16 or 8 bit words? Bits? The overhead of Python lists is not high compared to arrays, but the overhead of storing small quantities in large words can be substantial.

[AllanGErnst]

Thanks guys.

Peter: Yes I forgot to mention that its for storing 8 bit unsigned ints.

Regards Allan

[OutoftheBOTS_]

I did just have a little play with storing 8bit unsigned ints (unsigned Char) in an aaaray and this was sucessful but I wasn't able to do the same with 2D array.

MicroPython ESP32_LoBo_v3.1.28 - 2018-03-29 on ESP32 board with ESP32
Type "help()" for more information.
>>> from array import *
>>> my_array = array('B', [1,2,3,4,5])
>>> for item in my_array: print(item)
...
1
2
3
4
5
>>>

But this didn't work

>>> my_array = array('B', [[1,2,3,4,5],[1,2,3,4,5]])
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: can't convert list to int
>>>

[dhylands]

To store 8-bit ints you should probably use a bytearray. bytearrays are inherently one-dimensional, so I would create a wrapper class to make them look 2 dimensional and have it do the calculations. For example suppose I wanted a 2 dimensional array with 5 columns and 10 rows. You could allocate a 50-byte (5 * 10) entry byte-array, and given r and c calculate the index as r * 5 + c.

You could also use arrays but they're also inherently one dimensional, so they would need the same calculation.

You could also use a hybrid approach and create a 2 dimensional array by creating an array of byte arrays:

Code: Select all

>>> x = [bytearray(5), bytearray([1, 2, 3, 4, 5]), bytearray([6, 7, 8, 9, 0])]
>>> x
[bytearray(b'\x00\x00\x00\x00\x00'), bytearray(b'\x01\x02\x03\x04\x05'), bytearray(b'\x06\x07\x08\t\x00')]
>>> x[1][3]
4
>>> x[1][3] = 44
>>> x
[bytearray(b'\x00\x00\x00\x00\x00'), bytearray(b'\x01\x02\x03,\x05'), bytearray(b'\x06\x07\x08\t\x00')]
>>> x[1][3]
44
>>> ord(',')
44

The ASCII value of the comma in the second bytearray is 44. When printing bytearrays, it aways tries to use ASCII codes, and only uses the \x notation for non-ASCII values.

[OutoftheBOTS_]

@dhylands

I am a little confused by the output

>>> x[1][3] = 44
>>> x
[bytearray(b'\x00\x00\x00\x00\x00'), bytearray(b'\x01\x02\x03,\x05'), bytearray(b'\x06\x07\x08\t\x00')]

Why does position [1][3] seem to show 0x03 instead of 0x2c ????

[pythoncoder]

The contents of x[1] viewed as ordinal numbers are 1,2,3,44,5. So x[1][3] can be represented as 44 or 0x2c or ASCII ",".

[OutoftheBOTS_]

@pythoncoder
Ahh it wasn't till I looked close that I saw the "," that represents 44 or 0x2c

[pythoncoder]

We've all been there