Tracing an Uninitialized Error

ccelt09 has asked for the wisdom of the Perl Monks concerning the following question:

Hello Perl Monks! I am having difficulty discerning the source of an error in my code and or/input data. I attempt to filter data from one file to another based on what I call start and end positions.

 open (CG, "<$cg_input") or die "can't open $cg_input\n";
my @SNPs = <CG>;
close(CG); 

my $interval = "/Users/logancurtis-whitchurch/Desktop/chrX_divisions/"
+."$region"."_$filter".".txt"; #specifiecs intervals by region and fil
+ter version
open (INTERVAL, "<$interval") or die "can't open interval file\n";
foreach ( <INTERVAL> ) {
    
    my (undef, $start, $end) = split '\s+', $_;
    
    my $switch = 1; 
    while ($switch == 1) {         
    
        my @get_SNPs = split('\s+', $SNPs[$placeholder]);
        my $position = $get_SNPs[3];
                
        if (($position < $start) && ($position < $end)) {
            $placeholder++;
        }
        
        if (($position >= $start) && ($position <= $end)) {
            print OUT "@get_SNPs\n";
            $placeholder++;
        }
        if (($position > $start) && ($position > $end)) {
            $switch =! 1;    
        }
    }
}
close(INTERVAL);
[download]

The error message returned to me is as follows on repeating loop, until I terminate the program. Line 246 of my interval input file corresponds to the last line of the file, which looks no different from any of the others. I am at a loss regarding the uninitialized value, especially when identically formatted files have worked with the same code.

 Use of uninitialized value in split at filter.CGS.pl line 43, <INTERV
+AL> line 264.
Use of uninitialized value $position in numeric lt (<) at filter.CGS.p
+l line 46, <INTERVAL> line 264.
Use of uninitialized value $position in numeric lt (<) at filter.CGS.p
+l line 46, <INTERVAL> line 264.
Use of uninitialized value $position in numeric ge (>=) at filter.CGS.
+pl line 50, <INTERVAL> line 264.
Use of uninitialized value $position in numeric gt (>) at filter.CGS.p
+l line 54, <INTERVAL> line 264.
Use of uninitialized value in split at filter.CGS.pl line 43, <INTERVA
+L> line 264.
[download]

Comment on Tracing an Uninitialized Error Select or Download Code

Replies are listed 'Best First'.
Re: Tracing an Uninitialized Error by toolic (Bishop) on Jan 30, 2014 at 13:16 UTC
Tip #2 from the Basic debugging checklist: print `foreach ( <INTERVAL> ) { print;` [download]	[reply] [d/l]
Re^2: Tracing an Uninitialized Error by Anonymous Monk on Jan 30, 2014 at 21:27 UTC
Tip #5 from the Basic outright-wrong-code checklist: `while ( <INTERVAL> ) {` [download] It also reveals why 'line 264' seems to be the source of the errors.	[reply] [d/l]
Re: Tracing an Uninitialized Error by Random_Walk (Prior) on Jan 30, 2014 at 15:01 UTC
Print your ifs... `... print "if (($position < $start) && ($position < $end))\n"; if (($position < $start) && ($position < $end)) { ... # etc` [download] Cheers, R. Pereant, qui ante nos nostra dixerunt!	[reply] [d/l]
Re: Tracing an Uninitialized Error by GotToBTru (Prior) on Jan 30, 2014 at 15:20 UTC
Add the following to the beginning of your code, then run in debug: `$SIG{__WARN__} = sub { $DB::single = 1 };` That will stop execution whenever a warning is generated. You can inspect the variables at that point to determine which is undefined. Update: added = between } and sub	[reply] [d/l]
Re: Tracing an Uninitialized Error by Crackers2 (Parson) on Jan 30, 2014 at 17:21 UTC
Most likely explanation is that there's actually an empty line at the end of your data file, which in turn causes the last emelent of `@SNPs` to be an empty string, which in turn will cause the undefined `$position` Depending on your editor, the empty line might not be easy to spot. Doing a hexdump or similar might help.	[reply] [d/l] [select]
Re^2: Tracing an Uninitialized Error by ccelt09 (Sexton) on Jan 31, 2014 at 01:24 UTC
The last two lines of a hexdump looked like this for my input file, not quite an empty line but definitely asymmetrical / an incomplete line at the very end. `00019a0 30 35 36 30 09 31 35 35 32 37 30 35 36 30 00019ae` [download] I don't quite know what to make of this, if it is the source of the error or if it can be rectified	[reply] [d/l]
Re^3: Tracing an Uninitialized Error by Anonymous Monk on Jan 31, 2014 at 08:18 UTC
The problem probably is not on line 264. See 1072739.	[reply]