Parsing a numpy list read from binary via fromfile

Question

I have a 1D array of integers read in from a binary via

data = np.fromfile(file.dat, 'u1')

where data returns

array([242, 255, 101, 0, 124, 0, 86, 0, 33, 0, 
       112, 0, 32, 1, 74, 1, 195, 0, 16, 0], dtype=uint8)

The 0's in the series represent extra bits used for values over or under the standard 256. For example the 244 at index 0 should be -14 and is represented by 255 (or 256) - 242. Conversely the value at index 13 - reading as 32 is actuality 256 + 32, or 288. I'm trying to get the above array to look like --

[-14, 101, 124, 86, 33, 112, 288, 330, 195, 16]

I think I want something that uses the index and the value to evaluate the preceding value, if values 1 or 255 are found.

I've started with the below, but lost the thread at the end of the day.

for i, a in enumerate(data):
    if a == 255:
       neg = i,a

If anyone can recommend a better way to handle binary values I'm open to learning about that as well. Help appreciated!

What does lost the thread at the end of the day. mean?

wwii
– wwii

2019-05-14 00:45:09 +00:00
Commented May 14, 2019 at 0:45 — wwii
– wwii, Commented May 14, 2019 at 0:45
Ha, it's a euphemism for being tired

Derek_P
– Derek_P

2019-05-14 01:42:54 +00:00
Commented May 14, 2019 at 1:42 — Derek_P
– Derek_P, Commented May 14, 2019 at 1:42

Paul Panzer · Accepted Answer · 2019-05-14 00:49:22Z

1

You can viewcast to '<i2'(< stands for little endian, i for signed integer and 2 for 2 bytes)

x = np.array([242, 255, 101, 0, 124, 0, 86, 0, 33, 0, 
       112, 0, 32, 1, 74, 1, 195, 0, 16, 0], dtype=np.uint8)

x.view('<i2')
# array([-14, 101, 124,  86,  33, 112, 288, 330, 195,  16], dtype=int16)

Or you can use '<i2' directly with fromfile.

edited May 14, 2019 at 0:49

answered May 14, 2019 at 0:38

Paul Panzer

53.3k3 gold badges59 silver badges103 bronze badges

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

Collectives™ on Stack Overflow

Parsing a numpy list read from binary via fromfile

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related