printf (awk,perl,shell) float rounding issue


 
Thread Tools Search this Thread
Top Forums Shell Programming and Scripting printf (awk,perl,shell) float rounding issue
# 8  
Old 10-14-2012
Quote:
Originally Posted by Don Cragun
There are some statistics books that recommend rounding even values down (4.5 -> 4) and odd values up (5.5 -> 6) assuming that doing so will produce sums of rounded numbers closer to the value of the sums of the unrounded values. But I doubt that ksh is trying to provide that capability. I would guess that this could easily be caused by some programs using float, others using double, and still others using long double and having the process of converting 4.5 to the closest value in those formats sitting on opposite sides of the not quite exact value 4.5 representable in the various formats.
If it were something that is a random consquence of some conversion, then I think the rounding would not be as consistent as it is in those shells other than ksh... The rounding of half-integers are a special case and a choice needs to be made...

Apparently "round half to even" is the default and/or recommended way according to IEE754

Last edited by Scrutinizer; 10-14-2012 at 04:11 PM..
# 9  
Old 10-14-2012
Quote:
Originally Posted by Corona688
They must have taken effort to make it do so. I don't know why they bothered, since it's not a feature really needed, undesirable in most cases, and easy to do yourself if you really want it.

Floating point numbers are represented internally as an exponent, like 1.5*10^3 or so forth.
On the machines most of us work on, floating point numbers are represented as x*2^y or x*16^y rather than x*10^y, but .5 is exactly representable in any of them. On the other hand 0.1 is not exactly representable on any system that uses an exponent of 2 or 16.

As pointed out by Scrutinizer, the real answer lies in the rounding mode being used by whatever function is converting the internal floating point value to a printable decimal representation of that value.
# 10  
Old 10-14-2012
Quote:
Originally Posted by Don Cragun
On the machines most of us work on, floating point numbers are represented as x*2^y or x*16^y rather than x*10^y
I know how it works. It wasn't too relevant to the example. The point is it uses an exponent, it's not stored literally.
Quote:
but .5 is exactly representable in any of them.
0.5 is. Anything else .5 depends.
Quote:
As pointed out by Scrutinizer, the real answer lies in the rounding mode being used by whatever function is converting the internal floating point value to a printable decimal representation of that value.
Yes, which I pointed out most generally don't. ksh93 would be an exception.
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. Shell Programming and Scripting

[awk] rounding a float number?

Heyas Trying to calculate the total size of a file by reading its bitrate. Code snippet: fs_expected() { # # Returns the expected filesize in bytes # pr_str() { ff=$(cat $TMP.info) d="${ff#*bitrate: }" echo "${d%%,*}" | $AWK '{print $1}' | head -n 1 } t_BYTERATE=$((... (9 Replies)
Discussion started by: sea
9 Replies

2. UNIX for Dummies Questions & Answers

How to restrict Rounding using Printf?

Hello, I am using bash shell on Linux OS, May i please know why is it rounding for big numbers but not for others, is there a workaround to print it as it is with out round off? printf '%'\''.2f\n' 9999999999999999999.99 10,000,000,000,000,000,000.00 printf '%'\''.2f\n' 99999999999999.99... (1 Reply)
Discussion started by: Ariean
1 Replies

3. Shell Programming and Scripting

PERL - rounding fractional number

It seems that perl sprintf uses the round-to-even method: foreach my $i ( 0.5, 1.5, 2.5, 3.5 ) { printf "$i -> %.0f\n", $i; } __END__ 0.5 -> 0 1.5 -> 2 2.5 -> 2 3.5 -> 4 4.5 -> 4 Where we probably wants to use round-half-up, i.e. output should be as below: 0.5 -> 1 1.5 -> 2... (8 Replies)
Discussion started by: ganapati
8 Replies

4. Shell Programming and Scripting

AWK rounding up numbers

Hi, I have managed to round up numbers by using the following command: echo "5.54" | awk '{printf "%.0f\n", $1}' result 6 How can I round up all the numbers in a column in a file and print the lines with the new calculated totals? Thanks, (3 Replies)
Discussion started by: keenboy100
3 Replies

5. Shell Programming and Scripting

rounding time in perl

Is there a way I can round time in perl to the nearest five minutes? For example if I have log giving the following time stamps 23,52,30 it would rounded up to 23,55,00 and 23,50,01 would be rounded to 23,50,00 (3 Replies)
Discussion started by: borderblaster
3 Replies

6. Shell Programming and Scripting

Rounding issue with awk

Hi Friends, I am trying to round following number. 0.07435000 echo "0.07435000"|awk '{printf "%s\n",$1*100}'|awk '{printf "%.2f\n",$1}' It returns: 7.435 It should return: 7.44 Any suggestion please? Thanks, Prashant (2 Replies)
Discussion started by: ppat7046
2 Replies

7. Shell Programming and Scripting

How to format the output using float in awk{printf}

Hi I'm using awk to manipulate the data in the 6th field of the file xxx_yyy.hrv. The sample data that is available in this field is given below 220731.7100000000000000 When i tried using this command cat xxx_yyy.hrv | awk '{printf("%23.16f\n",$6*-1)}' I get the output as... (4 Replies)
Discussion started by: angelarosh
4 Replies

8. Shell Programming and Scripting

Rounding off decimals to the nearest number in PERL

Hi Guys, I am generating a statistical report , below is the snippet of the code : Now, $nSlices stands for the time duration,meaning,the statistics will be displayed for that particular time duration. Trouble is, for certain values of $totalTime (which is the end time - start time ), i... (9 Replies)
Discussion started by: rdlover
9 Replies

9. UNIX for Dummies Questions & Answers

Annoying rounding issue in awk

Hello I am getting this very annoying issue in awk: awk '{a=12825;b=a*1.25; print b}' test 16031.2 Thing is the multiplication result is wrong... Result should be 16031.25. I think the issue only happens on bigger numbers. What can I do to get passed this? Thanks by advance (3 Replies)
Discussion started by: Indalecio
3 Replies

10. Shell Programming and Scripting

Float and Double in Perl

HI! What is the notation which correspond to C's doubleandfloatin Perl?? Please tell me. Thanks. Well what I want to do is, for example, calculatingsqrt(2)in two way: float and double. (5 Replies)
Discussion started by: Euler04
5 Replies
Login or Register to Ask a Question