473,226 Members | 1,721 Online
Bytes | Software Development & Data Engineering Community
Post Job

Home Posts Topics Members FAQ

Join Bytes to post your question to a community of 473,226 software developers and data experts.

converting 64-bit fixed-point to float

Hi Group,

troubles with converting signed 32.32, little-endian, 2's complement
back to floating point. I have been trying to brew it myself. I am
running Python 2.5 on a Mac. Here is the C-code I have been trying to
leverage:

double FPuint8ArrayToFPDouble(uint8 *buffer, int startIndex)
{
uint32 resultDec = 0;
uint32 resultWh = 0;
int i;

for(i = 0; i < 4; i++)
{
resultDec += (uint32)buffer[startIndex + i] * pow(2, (i*8));
resultWh += (uint32)buffer[startIndex + i + 4] * pow(2, (i*8));
}

return ( (double)((int)resultWh) + (double)(resultDec)/4294967296.0 );
}
Here is my version in Python, with some test code built in:

from ctypes import *

def conv64(input):
input1=[0]*8
input1[0]=c_ushort(input[0])
input1[1]=c_ushort(input[1])
input1[2]=c_ushort(input[2])
input1[3]=c_ushort(input[3])
input1[4]=c_ushort(input[4])
input1[5]=c_ushort(input[5])
input1[6]=c_ushort(input[6])
input1[7]=c_ushort(input[7])
#print input1[0].value,
input1[1].value,input1[2].value,input1[3].value
#print
input1[4].value,input1[5].value,input1[6].value,input1[7].value
#print
resultDec=c_ulong(0)
resultWh=c_ulong(0)
for i in range(4):
dec_c=c_ulong(input1[i].value)
Wh_c=c_ulong(input1[i+4].value)
resultDec.value=resultDec.value+dec_c.value*2**(i* 8)
resultWh.value=resultWh.value+Wh_c.value*2**(i*8)
conval=float(int(resultWh.value))+float(resultDec. value)/4294967296.0
#print conval
return conval
#tabs got messed up bringing this into MacSoup

#these are 64-bit fixed point format (signed 32.32, little-endian, 2's
complement)
#should be -1
conv64_0=[0, 0, 0, 255, 255, 255, 255, 255]
#should be 0
conv64_1=[0, 0, 0, 0, 0, 0, 0, 0]
#should be 0.20000
conv64_1_2=[51, 51, 51, 51, 0, 0, 0, 0]
#should be 1
conv64_2=[0, 0, 0, 0, 1, 0, 0, 0]
#should be 2
conv64_3=[0, 0, 0, 0, 2, 0, 0, 0]
#should be 298.15
conv64_4=[102, 102, 102, 38, 42, 1, 0, 0]
#should be -0.2
conv64_5=[205,204,204,204,255,255,255,255]
output0=conv64(conv64_0)
print "output should be -1 is "+str(output0)
output1=conv64(conv64_1)
print "output should be 0 is "+str(output1)
output1_2=conv64(conv64_1_2)
print "output should be 0.2 is "+str(output1_2)
output2=conv64(conv64_2)
print "output should be 1 is "+str(output2)
output3=conv64(conv64_3)
print "output should be 2 is "+str(output3)
output4=conv64(conv64_4)
print "output should be 298.15 is "+str(output4)
output5=conv64(conv64_5)
print "output should be -0.2 is "+str(output5)

Finally, here is the output I get from my code:
>>>
output should be -1 is 4294967296.0
output should be 0 is 0.0
output should be 0.2 is 0.199999999953
output should be 1 is 1.0
output should be 2 is 2.0
output should be 298.15 is 298.15
output should be -0.2 is 4294967295.8

Thanks for any light you can shed on my ignorance.

wave_man
Jul 21 '07 #1
2 2865
It appears to be correct for positive numbers.

if conval >= 2**16:
conval -= 2**32

would appear to patch things up.

It's not very pretty, though. You could at least start with

input1 = [c_ushort(item) for item in input]

instead of your first 9 lines.

mt

Jul 21 '07 #2
On Jul 20, 5:59 pm, johnmfis...@comcast.net (John Fisher) wrote:
Hi Group,

troubles with converting signed 32.32, little-endian, 2's complement
back to floating point. I have been trying to brew it myself. I am
running Python 2.5 on a Mac. Here is the C-code I have been trying to
leverage:

double FPuint8ArrayToFPDouble(uint8 *buffer, int startIndex)
{
uint32 resultDec = 0;
uint32 resultWh = 0;
int i;

for(i = 0; i < 4; i++)
{
resultDec += (uint32)buffer[startIndex + i] * pow(2, (i*8));
resultWh += (uint32)buffer[startIndex + i + 4] * pow(2, (i*8));
}

return ( (double)((int)resultWh) + (double)(resultDec)/4294967296.0 );

}

There are a few problem spots in that C code. I tend to
think that it "works" because you're on a system that has
4-byte int and CHAR_BIT == 8. When the most-significant-bit (MSB)
of resultWh is 1, then casting to int makes that a negative
value (i.e., MSB == the sign bit).

I presume that somewhere you include <stdint.h(from C99)
and that uint32 is really uint32_t, etc. For that to be
portable, you should probably cast to int32_t?

#include <stdio.h>
#include <stdlib.h>
#include <stdint.h>

double arr2dbl (uint8_t *buffer, int startIndex)
{
uint32_t decimal = 0;
uint32_t whole = 0;
size_t i;
for (i = 0; i < 4; ++i)
{
decimal += (buffer[startIndex + i] << (i*8));
whole += (buffer[startIndex + i + 4] << (i*8));
}
return (int32_t)whole + (decimal/(UINT32_MAX+1.0));
}

int main (void)
{
uint8_t arr[7][8] = {
{0, 0, 0, 0, 0xff, 0xff, 0xff, 0xff},
{0, 0, 0, 0, 0, 0, 0, 0},
{51, 51, 51, 51, 0, 0, 0, 0},
{0, 0, 0, 0, 1, 0, 0, 0},
{0, 0, 0, 0, 2, 0, 0, 0},
{102, 102, 102, 38, 42, 1, 0, 0 },
{205, 204, 204, 204, 0xff, 0xff, 0xff, 0xff}};
size_t i;
double result;
for (i = 0; i < sizeof arr/sizeof arr[0]; ++i)
{
result = arr2dbl(arr[i], 0);
printf("%f\n", result);
}
return 0;
}


Here is my version in Python, with some test code built in:

from ctypes import *

def conv64(input):
input1=[0]*8
input1[0]=c_ushort(input[0])
input1[1]=c_ushort(input[1])
input1[2]=c_ushort(input[2])
input1[3]=c_ushort(input[3])
input1[4]=c_ushort(input[4])
input1[5]=c_ushort(input[5])
input1[6]=c_ushort(input[6])
input1[7]=c_ushort(input[7])
#print input1[0].value,
input1[1].value,input1[2].value,input1[3].value
#print
input1[4].value,input1[5].value,input1[6].value,input1[7].value
#print
resultDec=c_ulong(0)
resultWh=c_ulong(0)
for i in range(4):
dec_c=c_ulong(input1[i].value)
Wh_c=c_ulong(input1[i+4].value)
resultDec.value=resultDec.value+dec_c.value*2**(i* 8)
resultWh.value=resultWh.value+Wh_c.value*2**(i*8)
conval=float(int(resultWh.value))+float(resultDec. value)/4294967296.0
#print conval
return conval
#tabs got messed up bringing this into MacSoup

(snipped)

>
Finally, here is the output I get from my code:

output should be -1 is 4294967296.0
output should be 0 is 0.0
output should be 0.2 is 0.199999999953
output should be 1 is 1.0
output should be 2 is 2.0
output should be 298.15 is 298.15
output should be -0.2 is 4294967295.8

Thanks for any light you can shed on my ignorance.

wave_man

This is my translation:

from ctypes import *

def conv64(input):
input1 = [c_uint8(item) for item in input]
dec = c_uint32(0)
whl = c_uint32(0)
for i in xrange(4):
dec.value += (input1[i].value << (i*8))
whl.value += (input1[i+4].value << (i*8))
cast_whl_to_int = c_int32(whl.value)
return float(cast_whl_to_int.value + dec.value/4294967296.0)
for arr in [[0, 0, 0, 0, 0xff, 0xff, 0xff, 0xff],
[0, 0, 0, 0, 0, 0, 0, 0],
[51, 51, 51, 51, 0, 0, 0, 0],
[0, 0, 0, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 2, 0, 0, 0],
[102, 102, 102, 38, 42, 1, 0, 0],
[205,204,204,204,255,255,255,255]]:
print "%f" % conv64(arr)

However, I've not looked deeply into ctypes so I
don't know if c_int32 is really C's int, or int32_t, or ???

--
Hope this helps,
Steven

Jul 21 '07 #3

This thread has been closed and replies have been disabled. Please start a new discussion.

Similar topics

11
by: Chris Online | last post by:
Hi all, I'm using C++ Builder5. I want to get data from an edit-box and send it to a development kit. The dev-kit can only receive char and no char* here's a part of my code: char* Data_byte...
7
by: Jus! | last post by:
Hi. I am reading bits(1's & 0's) from a file and i wa wondering what is the most efficient method of converting these strings to individual int's? eg. File contains: 110001 010011 etc......
8
by: Ramiro Barbosa, Jr. | last post by:
All, Any ideas on how to convert the first 8 bytes of raw uninterpreted sequence of bytes from 'char array;' (populated with _binary_ data read from a socket), into a 'long id'? Thank you! ...
3
by: j.a. harriman | last post by:
Hi, On MSDN I know there is a JScript example (Upgrading Visual C++ Projects to Visual Studio .NET in Batch Mode) to upgrade VS6 C++ projects to .NET solutions. It converts the project files...
19
by: silentlights | last post by:
Hi, Can you help me conver a char array into an integer. I am trying something like.. char carray; int numb; carray = 1; carray = 5; carray = 1;
1
by: scott | last post by:
Hi all, trying to use base64. Ill get right to the problem. I am converting a string into base 64. No problem there. That base64 string can then be converted back to the orignal string. No...
8
by: moondaddy | last post by:
I need to convert a byte array to a string and pass it as a parameter in a URL and then convert it back to the original byte array. However, its getting scrambled in the conversion. In short,...
5
by: byte8bits | last post by:
Here's how I'm doing this right now, It's a bit slow. I've just got the code working. I was wondering if there is a more efficient way of doing this... simple example from interactive Python: ...
2
by: cablepuff | last post by:
template <typename ContainerType> ContainerType rsa_encrypt_list(const std::string&, const typename ContainerType::reference, const typename ContainerType::reference); const BigInteger...
2
by: joe shoemaker | last post by:
I would like to convert url into md5 hash. My question is that md5 hash will create collision at 2^64. If you do long(value,16), where value is the md5 hash string, would value returned from...
0
by: veera ravala | last post by:
ServiceNow is a powerful cloud-based platform that offers a wide range of services to help organizations manage their workflows, operations, and IT services more efficiently. At its core, ServiceNow...
0
by: VivesProcSPL | last post by:
Obviously, one of the original purposes of SQL is to make data query processing easy. The language uses many English-like terms and syntax in an effort to make it easy to learn, particularly for...
3
isladogs
by: isladogs | last post by:
The next Access Europe meeting will be on Wednesday 3 Jan 2024 starting at 18:00 UK time (6PM UTC) and finishing at about 19:15 (7.15PM). For other local times, please check World Time Buddy In...
0
by: mar23 | last post by:
Here's the situation. I have a form called frmDiceInventory with subform called subfrmDice. The subform's control source is linked to a query called qryDiceInventory. I've been trying to pick up the...
2
by: jimatqsi | last post by:
The boss wants the word "CONFIDENTIAL" overlaying certain reports. He wants it large, slanted across the page, on every page, very light gray, outlined letters, not block letters. I thought Word Art...
2
isladogs
by: isladogs | last post by:
The next Access Europe meeting will be on Wednesday 7 Feb 2024 starting at 18:00 UK time (6PM UTC) and finishing at about 19:30 (7.30PM). In this month's session, the creator of the excellent VBE...
0
by: fareedcanada | last post by:
Hello I am trying to split number on their count. suppose i have 121314151617 (12cnt) then number should be split like 12,13,14,15,16,17 and if 11314151617 (11cnt) then should be split like...
0
Git
by: egorbl4 | last post by:
Скачал я git, хотел начать настройку, а там вылезло вот это Что это? Что мне с этим делать? ...
0
by: MeoLessi9 | last post by:
I have VirtualBox installed on Windows 11 and now I would like to install Kali on a virtual machine. However, on the official website, I see two options: "Installer images" and "Virtual machines"....

By using Bytes.com and it's services, you agree to our Privacy Policy and Terms of Use.

To disable or enable advertisements and analytics tracking please visit the manage ads & tracking page.