Question

I have a large binary file of ieee 32bit floating point numbers.

In python I use:

f = file.read(4)
    while f !='':
        if len(f) == 4:
            data =struct.unpack('>f', f)
            print data
f = file.read(4)

to read it 4 bytes at a time

However, occasionally f would be size 1, and struct.unpack would complain that its input must be a string of size 4.

The filesize is divisible by 4, and this happens multiple times within the file.

What could be causing this?

Answer 1

Did you open the file in binary mode?

Anyway, a much better way to read your file is to use array.fromfile() or NumPy.