I dont wether I got u right. I think this is what you are looking at

#!/usr/bin/perl
#
# Read the entire Data into a single string
#

$/=undef;
$string =  <DATA>;
$string=~s/\s+/ /g;
$prev = 0;
while($string=~/(\S+) (\S+) (\S+)/g) {
     print "$2 =>($3 - $prev)  " . ( $3 - $prev ) . " \n";
     $prev = $3;
}

__DATA__
1.40616  tw-
1.51162
1.53491  w-ih
1.55572
1.61759  ihLx
1.70433
1.74422  Lx#
1.91998
1.94224  UU
1.95236




Dennis Warren wrote:
> Hi,
> 
> I'm trying to use data to work out the average durations of certain sounds
> for a project here at work. The problem I have is that I'm a real newbie at
> Perl, and I'm finding the data really hard to get my head around. I don't
> really know where to begin. Any suggestions for an approach would be really
> appreciated.
> 
> The data, in plain text files, looks something like this:
> 
>     1.40616  tw-
>     1.51162 
>     1.53491  w-ih
>     1.55572 
>     1.61759  ihLx
>     1.70433 
>     1.74422  Lx#
>     1.91998
>     1.94224  UU
>     1.95236  
> 
> The first column is a time in seconds, and the second column (not always
> filled) represents the symbol/sound I'm measuring.
>    
> What I need to do is to work out the length of a certain symbol. (the one in
> the 2nd column, e.g. "w-ih"). I have to do this by taking the time in the
> previous record away from the time in the next record to get the duration
> 
>     1.40616  tw-
>     1.51162           <- the start time of the symbol to measure
>     1.53491  w-ih             <- the symbol to measure
> e.g. 1.5572 - 1.51162 = 0.04558    
>     1.55572           <- the stop time of the symbol to measure
>     1.61759  ihLx
>     1.70433 
> 
> As you can see, the stop time for one sound is very often also the start
> time for the next sound. As well as this I face the problem that the first
> sound in all the records cannot be measured, as it has no start point. So I
> must always ignore this. I want to output the durations of all the symbols
> in a given data set, e.g..
> 
> w-ih  0.04558
> ihLx  0.14861
> 
> 
> Any solutions, or hints at achieving a solution would be great
> 
> regards,
> 
> Dennis


-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to