On Fri, Jan 25, 2013 at 4:13 AM, IZ3GME Marco <iz3gme.marco@gmail.com> wrote:

It's a rounding problem due to the internal representation of float.

I tried the following version of convert_freq_to_bytes which seems to
works but may be Dan or some other python expert have a better solution.

def convert_freq_to_bytes(freq):

bytes = [ 0 for x in range(0,8) ] # init list with 8 times 0

real_freq = "%3.5f" % freq # now is a string
real_freq = int(real_freq[0:3]+ real_freq[4:])
# and now an int

for i in range(7, -1, -1): # go from 7 to 0
bytes[i] = real_freq%10 # extract last digit
real_freq /= 10 # throw away last digit
return bytes

73 de IZ3GME Marco

I still get 146.63999 out for 146.64000 in.

In settings.py, I had changed 'precision=4' to 'precision=5'

class RadioSettingValueFloat(RadioSettingValue):
"""A floating-point setting"""
def __init__(self, minval, maxval, current, resolution=0.001, precision=4):

The reason I did this was to get 5 places past the decimal. Maybe that was the wrong way to accomplish that. Anyway, I changed back to 'precision=4' and now the result is 146.6400, which is better than 146.63999, but still not exactly what I am looking for.

Thanks again. You help and guidance is greatly appreciated.
Jim KC9HI