#!/usr/bin/perl

print "gnuplotme v0.6 by hani jamjoom\n";

$usage =  " see http://jamjoom.net/tools/gnuplotme for complete info.\n\n ".
          " gnuplotme --help\n" .
	  "                 --outfile <file_base_name> [eps | jpeg | png]\n" . 
	  "                 --file <file_name> <plot_label> <column>\n" .
	  "                 --xlabel <x_axes_label>\n" .
	  "                 --ylabel <y_axes_label> [<axis>]\n" .	      
	  "                 --xrange <low:hi> \n" .
	  "                 --yrange <low:hi> [<axis>]\n" .
	  "                 --title <title>\n" .
	  "                 --key <x,y>\n" .    
	  "                 --percent  [left,right,both]\n" .
	  "                 --log_scale <x or y or xy>\n\n" .
	  "                 --subsample <period>\n\n" .
	  "           advanced options:\n" .
	  "                 --set  <variable_name> { {file  <filename> <column>} |\n" .
	  "                                          {range low high step} |\n" .
	  "                                          {list  y1 y2 y3 ... } |\n" .
          "                                          {avg   <column> <outliers> <conf> file1 file2 ... }}\n".
	  "                 --plot  <function> <plot_label> [<axis>]\n" .
	  "                 --ytics <low,step,hi> [<axis>]\n\n";

# Table for T-Distribution.  The last element of each entry is for very large (infinity)
# population sizes.

%tdist = ( 20 => [ .325, .289, .277, .271, .267, .265, .263, .262, .261, .260, 
	      .260, .259, .259, .258, .258, .258, .257, .257, .257, .257, 
	      .257, .256, .256, .256, .256, .256, .256, .256, .256, .256, 
	      .253 ],
	   40 => [ .727, .617, .584, .569, .559, .553, .549, .546, .543, .542,
	      .540, .539, .538, .537, .536, .535, .534, .534, .533, .533,
	      .532, .532, .532, .531, .531, .531, .531, .530, .530, .530,
	      .524 ],
	   60 => [ 1.367, 1.061, .978, .941, .920, .906, .896, .889, .883, .879,
	      .876, .873, .870, .868, .866, .865, .863, .862, .861, .860,
	      .859, .858, .858, .857, .856, .856, .855, .855, .854, .854,
	      .842 ],
	   80 => [ 3.078, 1.886, 1.638, 1.533, 1.476, 1.440, 1.415, 1.397, 1.383, 1.372,
	      1.363, 1.356, 1.350, 1.345, 1.341, 1.337, 1.333, 1.330, 1.328, 1.325,
	      1.323, 1.321, 1.319, 1.316, 1.316, 1.315, 1.314, 1.313, 1.310, 1.310,
	      1.282 ],
	   90 => [ 6.314, 2.920, 2.353, 2.132, 2.015, 1.943, 1.895, 1.860, 1.833, 1.812,
	      1.796, 1.782, 1.771, 1.761, 1.753, 1.746, 1.740, 1.734, 1.729, 1.725,
	      1.721, 1.717, 1.714, 1.708, 1.708, 1.706, 1.703, 1.701, 1.697, 1.697,
	      1.645 ],
	   95 => [ 12.706, 4.303, 3.182, 2.776, 2.571, 2.447, 2.365, 2.306, 2.262, 2.228,
	      2.201, 2.179, 2.160, 2.145, 2.131, 2.120, 2.110, 2.101, 2.093, 2.086,
	      2.080, 2.074, 2.069, 2.060, 2.060, 2.056, 2.052, 2.048, 2.042, 2.042,
	      1.960 ],
	   98 => [ 31.821, 6.965, 4.541, 3.747, 3.365, 3.143, 2.998, 2.896, 2.821, 2.764,
	      2.718, 2.681, 2.650, 2.624, 2.602, 2.583, 2.567, 2.552, 2.539, 2.528,
	      2.518, 2.508, 2.500, 2.485, 2.485, 2.479, 2.473, 2.467, 2.457, 2.457,
	      2.326 ],
	   99 => [ 63.657, 9.925, 5.841, 4.604, 4.032, 3.707, 3.499, 3.355, 3.250, 3.169,
	      3.106, 3.055, 3.012, 2.977, 2.947, 2.921, 2.898, 2.878, 2.861, 2.845,
	      2.831, 2.819, 2.807, 2.787, 2.787, 2.779, 2.771, 2.763, 2.750, 2.750,
	      2.576 ]  );

$tdist_infinity = 30;

sub numerically { $a <=> $b }

if ( $#ARGV < 0 ) {
    print "USAGE: $usage";

    exit(0);
}

# PARSE OPTIONS

$log_scale = 0;
$i=0;
$j=0;
$subsample = 1 ;
$var_cnt_min = 9999;
$func_cnt = 0;
$file_cnt = 0;

while ( $#ARGV >= 0 ) {

    $in_arg = shift(@ARGV);


    if ( $in_arg eq "--help" ) {

	print "$usage";
	exit(0);

    } elsif ( $in_arg eq "--log_scale" ) {
	
	$log_scale = shift(@ARGV);

	if ( ($log_scale ne "x") && ($log_scale ne "y") && ($log_scale ne "xy") ) {
	    print "SYNTAX ERROR: --log_scale option must either be x, y, or xy.\n";
	    exit(0);	    
	} 

    } elsif ( $in_arg eq "--subsample" ) {
	
	$subsample = shift(@ARGV);

    } elsif ( $in_arg eq "--confidence" ) {
	
	$subsample = shift(@ARGV);

    } elsif ( $in_arg eq "--percent" ) {
	
	$percent = shift(@ARGV);

	if ( ($percent ne "left") && ($percent ne "right") && ($percent ne "both") ) {
	    print "SYNTAX ERROR: --percent option must either be left, right, or both.\n";
	    exit(0);	    
	} 

    } elsif ( $in_arg eq "--xlabel" ) {

	$xlabel  = shift(@ARGV);

    } elsif ( $in_arg eq "--xrange" ) {

	$x_range  = shift(@ARGV);

    } elsif ( $in_arg eq "--title" ) {

	$title  = shift(@ARGV);

    } elsif ( $in_arg eq "--ytics" ) {

	$tmp1 = shift(@ARGV);
	$tmp2 = shift(@ARGV);
	
 	if ($tmp2 ne "" ) {
	    if (($tmp2 eq "left") || ($tmp2 eq "right")) {
	    
		if ($tmp2 eq "left") {
		    $y_tics  = $tmp1;
		}

		if ($tmp2 eq "right") {
		    $y_tics_right  = $tmp1;
		}

	    } elsif ( index($tmp2,"--") == -1 ) {
		print "SYNTAX ERROR: --ytics second option must either be left or right.\n";
		exit(0);	    
	    } else {
		$y_tics  = $tmp1;
		unshift(@ARGV,$tmp2);
	    }

	} else {
	    $y_tics  = $tmp1;
	}

    } elsif ( $in_arg eq "--yrange" ) {

	$tmp1 = shift(@ARGV);
	$tmp2 = shift(@ARGV);
	
 	if ($tmp2 ne "" ) {
	    if (($tmp2 eq "left") || ($tmp2 eq "right")) {
	    
		if ($tmp2 eq "left") {
		    $y_range  = $tmp1;
		}

		if ($tmp2 eq "right") {
		    $y_range_right  = $tmp1;
		}

	    } elsif ( index($tmp2,"--") == -1 ) {
		print "SYNTAX ERROR: --yrange second option must either be left or right.\n";
		exit(0);	    
	    } else {
		$y_range  = $tmp1;
		unshift(@ARGV,$tmp2);
	    }

	} else {
	    $y_range  = $tmp1;
	}

    } elsif ( $in_arg eq "--ylabel" ) {

	$tmp1 = shift(@ARGV);
	$tmp2 = shift(@ARGV);

	if ($tmp2 ne "" ) {
	    if (($tmp2 eq "left") || ($tmp2 eq "right")) {
	    
		if ($tmp2 eq "left") {
		    $ylabel  = $tmp1;
		}

		if ($tmp2 eq "right") {
		    $ylabel_right  = $tmp1;
		}

	    } elsif ( index($tmp2,"--") == -1 ) {
		print "SYNTAX ERROR: --ylabel second option must either be left or right.\n";
		exit(0);	    
	    } else {
		$ylabel  = $tmp1;
		unshift(@ARGV,$tmp2);
	    }

	} else {
	    $ylabel  = $tmp1;
	}


    } elsif ( $in_arg eq "--key" ) {

	$key  = shift(@ARGV);

   } elsif ( $in_arg eq "--outfile" ) {

	$outfile  = shift(@ARGV);
	$outtype  = shift(@ARGV);

	if ($outtype ne "" ) {

	    if (($outtype eq "eps") || 
		($outtype eq "png") ||
		($outtype eq "jpeg") ) {

		#noop

	    } elsif ( index($outtype,"--") == -1 ) {
		print "SYNTAX ERROR: --outtype second option must be a valid type.\n";
		exit(0);	    
	    } else {
		unshift(@ARGV,$outtype);
		$outtype  = "eps";
	    }

	} else {
	    $outtype  = "eps";
	}

    } elsif ( $in_arg eq "--file" ) {

# Both of the following are following:
#        --file <file_name> <plot_label> <column>
# or 
#        --set f_n file <file_name> <column>
#        --plot f_n label

	$in_name   = shift(@ARGV);
	$in_label  = shift(@ARGV);
	$in_column = shift(@ARGV);

	if ( ($in_name eq "")   || (index($in_name,"--")  != -1)  || 
	     ($in_label eq "")  || (index($in_label,"--") != -1)  ||
	     ($in_column eq "") || (index($in_column,"--")!= -1))  {

	    print "SYNTAX ERROR: --file ($in_name) is missing something dude!!\n";
	    exit(0);
	}

	splice(@ARGV,0,0,"--set","F_$file_cnt","file","$in_name","$in_column");
	splice(@ARGV,0,0,"--plot","F_$file_cnt","$in_label");

	$file_cnt++;

    }  elsif ( $in_arg eq "--set" ) {
	
	$in_var = shift(@ARGV);
	$in_type = shift(@ARGV);

	if ( $in_type eq "file" ) {
	    
	    $in_name = shift(@ARGV);
	    $in_column = shift(@ARGV) - 1;

	    open(FD,"$in_name");
	    
	    $cnt = 0;
	    while(<FD>) {
		
		@line = split(/\s+/,$_);

		$var{$in_var}[$cnt]=$line[$in_column];
		
		$cnt++;
	    }

	    $var_cnt{$in_var} = $cnt;
	    
	    close(FD);
	    
	} elsif ( $in_type eq "range" ) {
	    
	    $in_low = shift(@ARGV);
	    $in_high = shift(@ARGV);
	    $in_step = shift(@ARGV); 

	    $cnt=0;
	    for($i=$in_low; $i<=$in_high; $i+=$in_step) {
		
		$var{$in_var}[$cnt] = $i;
		$cnt++;
	    }
	    
	    $var_cnt{$in_var} = $cnt;

	} elsif ( $in_type eq "list" ) {

	    $cnt=0;
	    $in_val = shift(@ARGV);

	    while ( (index($in_val,"--") == -1 ) && ($in_val ne "") ) {
		$var{$in_var}[$cnt] = $in_val;
		$cnt++;
		$in_val = shift(@ARGV);
	    }
	    
	    $var_cnt{$in_var} = $cnt;

	    if ( index($in_val,"--") != -1 ) {
		unshift(@ARGV,$in_val);
	    }

	} elsif ( $in_type eq "avg" ) {

	    $in_column  = shift(@ARGV) - 1;
	    $in_outlier = shift(@ARGV);
	    $in_conf    = shift(@ARGV);
	    $in_name    = shift(@ARGV);

	    $fcnt=0;
	    $min_rcnt = 99999;

	    while ( (index($in_name,"--") == -1 ) && ($in_name ne "") ) {

		open(FD,"$in_name");
	    
		print "$in_name\n";

		$rcnt = 0;
		while(<FD>) {

		    @line = split(/\s+/,$_);

		    $in_avg[$fcnt][$rcnt]=$line[$in_column];
		    $rcnt++;
		}
		
		close(FD);
		
		if ($rcnt < $min_rcnt) {
		    $min_rcnt = $rcnt;
		}

		$in_name = shift(@ARGV); 

		$fcnt++;
	    }
	    
	    if ( index($in_name,"--") != -1 ) {
		unshift(@ARGV,$in_name);
	    }

	    $var_cnt{$in_var} = $min_rcnt;

	    # compute the mean and confidence of the columns 

	    if ( $fcnt != 0 ) {
		
		for ($j=0;$j<$min_rcnt;$j++) {
		    
		    #translate columns into rows
		    
		    for ($i=0;$i<$fcnt;$i++) {
			$tmp_c[$i] = $in_avg[$i][$j];
		    }

		    @sorted_tmp_c = sort numerically @tmp_c; 

		    $tmp = 0;
		    for ($i=$in_outlier;$i<$fcnt-$in_outlier;$i++) {
			$tmp += $sorted_tmp_c[$i];
		    }

		    $population = $fcnt-2*$in_outlier;
		    if ( $population > 0 ) {
			$tmpmean = $tmp/$population; 
		    } else {
			$tmpmean = 0;
		    }

		    $tmpvar = 0;
		    for ($i=$in_outlier;$i<$fcnt-$in_outlier;$i++) {
			$tmpvar += ($sorted_tmp_c[$i]-$tmpmean)**2;
		    }
		
		    $var{$in_var}[$j] = $tmpmean;		    

		    if ( (($in_conf == 20) ||
			 ($in_conf == 40) ||
			 ($in_conf == 60) ||
			 ($in_conf == 80) ||
			 ($in_conf == 90) ||
			 ($in_conf == 95) ||
			 ($in_conf == 98) ||
			 ($in_conf == 99)) && 
			 ($population > 0) ) {

			if ( ($population-1) < $tdist_infinity ) {
			    $tval = $tdist{$in_conf}[$population-1];
			} else {
			    $tval = $tdist_infinity;
			}
			
			$conf{$in_var}[$j] = $tval*sqrt($tmpvar)/sqrt($population);
		    }
		    
		    undef @tmp_c;
		    undef @sorted_tmp_c;
		}
	    }

	    undef @in_avg;

	} else {
	    print "SYNTAX ERROR: You wrote something that is not supported ($in_type)\n";
	}
	
    } elsif ( $in_arg eq "--plot" ) {

	$in_func   = shift(@ARGV);
	$in_label  = shift(@ARGV);
	$in_axes = shift(@ARGV);

	$in_func =~ s/\s//g; #remove spaces 
	
	$func[$func_cnt] = $in_func;  
	$func_label[$func_cnt] = $in_label;

	if ( $in_axes ne "" ) {
	    if (($in_axes eq "left") || ($in_axes eq "right")) {
		$func_axes[$func_cnt] = $in_axes;
	    } elsif (index($in_axes,"--")==-1) {
		print "SYNTAX ERROR: --func third option ($in_axes) must either be left or right.\n";
		exit(0);
	    } else {
		$func_axes[$func_cnt] = "left";
		unshift(@ARGV,$in_axes);
	    }
	}

	$func_cnt++;
	
    } else {
	print "SYNTAX ERROR: You wrote something that is not supported ($in_arg)\n";
	exit(0);
    }
}


# make sure that all vars have the same number of entries
# if not find the minumum

foreach $key (sort keys(%var_cnt)) {
    if ( $var_cnt{$key} < $var_cnt_min ) {
	$var_cnt_min = $var_cnt{$key};
    }
}

foreach $key (sort keys(%var_cnt)) {

    if ( $var_cnt{$key} > $var_cnt_min ) {
	print "WARNING: variable ($key) is longer than the minimum of all variables\n";
	print "         will truncate its length to $var_cnt_min\n";
	$var_cnt{$key} = $var_cnt_min;
    }
}

# compute mean, total, variance

foreach $key (sort keys(%var)) {

    for ($i = 0; $i < $var_cnt{$key}; $i++) {
	$var_total{$key} += $var{$key}[$i];
    }

    if ($var_cnt{$key} != 0) {

	$var_mean{$key} = $var_total{$key}/$var_cnt{$key};
	
	for ($i = 0; $i < $var_cnt{$key}; $i++) {
	    $var_variance{$key} += ($var{$key}[$i]-$var_mean{$key})**2;
	}
    }

#    print "[$key]->";
#    for ($i = 0; $i < $var_cnt{$key}; $i++) {
#	print "($var{$key}[$i])->";
#    }
#    print "[total=$var_total{$key},mean=$var_mean{$key},variance=$var_variance{$key}]\n";

}

if (!$outfile) {
    print "SYNTAX ERROR: --outfile option must be specified\n";
    exit(0);
}


$outdata = $outfile . ".data";
open (FD,"> $outdata");

for ($i=0; $i<$var_cnt_min; $i += $subsample) {

    for ($j=0; $j<=$#func; $j++) {

	$tmpfunc = $func[$j]; 

	foreach $key (sort keys(%var)) {

	    # first replace the mean, variance, count, and total

	    $tmpfunc =~ s/mean_$key/$var_mean{$key}/gi;
	    $tmpfunc =~ s/var_$key/$var_variance{$key}/g;
	    $tmpfunc =~ s/count_$key/$var_cnt{$key}/g;
	    $tmpfunc =~ s/total_$key/$var_total{$key}/g;

	    # now replace all instances
	    $tmpfunc =~ s/$key/$var{$key}[$i]/g;
	}

	$tmp = eval($tmpfunc);
	
	if ( defined($conf{$func[$j]})  ) {
	    printf(FD "%0.3f\t%0.3f\t",$tmp,$conf{$func[$j]}[$i]);
	} else {
	    printf(FD "%0.3f\t",$tmp);
	}
    }

    print FD "\n";
}

close(FD);

# Lets do some plotting 

$plot_command = 'plot ';

if ($func_label[0] eq "xaxis") {
    $j=1;
} else {
    $j=0;
}
 
for ($i=$j; $i <= $#func; $i++ ) {

    $column = $i+1;

    if ( $func_axes[$i] eq "right" ) {
	$tmpaxes = "x1y2";
    } else {
	$tmpaxes = "x1y1";
    }

    if ( defined($conf{$func[$j]})  ) {
	$cp1 = $column+1;
	$plot_command = $plot_command . "\"$outdata\" using 1:$column:$cp1 axes $tmpaxes" .
	    " title \"$func_label[$i]\" with yerrorbars";
	$i++;
    } else {
	$plot_command = $plot_command . "\"$outdata\" using 1:$column axes $tmpaxes" .
	    " title \"$func_label[$i]\" with linespoint";
    }
    
    if ( $i < $#func ) {
	$plot_command = $plot_command . ', ';
    }
}

open ( PLOT_FILE, "> $outfile.gnu");
	
#print PLOT_FILE  "set terminal postscript eps enhanced \"Helvetica\" 20 \n";
if ( $outtype eq "jpeg" ) {
    print PLOT_FILE  "set terminal jpeg\n";
    print PLOT_FILE  "set output \"$outfile.jpg\"\n";
} elsif ( $outtype eq "png" ) {
    print PLOT_FILE  "set terminal png\n";
    print PLOT_FILE  "set output \"$outfile.png\"\n";
} else { #default to eps
    print PLOT_FILE  "set terminal postscript eps\n";
    print PLOT_FILE  "set output \"$outfile.eps\"\n";
}

print PLOT_FILE  "set xlabel \"$xlabel\"\n";
print PLOT_FILE  "set ylabel \"$ylabel\"\n";
#print PLOT_FILE  "set size .6, .6\n";
print PLOT_FILE  "set title \"$title\"\n";

if ($key ne "" ) {
    print PLOT_FILE  "set key $key\n";
}

if ($x_range ne "" ) {
    print PLOT_FILE  "set xrange [$x_range]\n";
}

if ($y_range ne "" ) {
    print PLOT_FILE  "set yrange [$y_range]\n";
}

if ($y_range_right ne "" ) {
    print PLOT_FILE  "set y2range [$y_range_right]\n";
}

if ($y_tics ne "" ) {
    print PLOT_FILE  "set ytics $y_tics\n";
}

if ($y_tics_right ne "" ) {
    print PLOT_FILE  "set y2tics $y_tics_right\n";
}

if ($ylabel_right ne "" ) {
    print PLOT_FILE  "set y2label \"$ylabel_right\"\n";
}

if ($percent eq "right" ) {
    print PLOT_FILE  "set y2tics  0,25,100\n";
    print PLOT_FILE  "set y2range [0:100]\n";
}

if ($percent eq "left" ) {
    print PLOT_FILE  "set ytics   0,25,100\n";
    print PLOT_FILE  "set yrange  [0:100]\n";
}

if ($percent eq "both" ) {
    print PLOT_FILE  "set y2tics  0,25,100\n";
    print PLOT_FILE  "set ytics   0,25,100\n";
    print PLOT_FILE  "set yrange  [0:100]\n";
    print PLOT_FILE  "set y2range [0:100]\n";
}

if ($log_scale eq "x" ) {
    print PLOT_FILE  "set logscale x\n";
}

if ($log_scale eq "y" ) {
    print PLOT_FILE  "set logscale y\n";
}

if ($log_scale eq "xy" ) {
    print PLOT_FILE  "set logscale x\n";
    print PLOT_FILE  "set logscale y\n";
}

print PLOT_FILE  "$plot_command\n";

close( PLOT_FILE);

`gnuplot $outfile.gnu`;