/usr/bin/gpiv_t-avg is in gpivtools-mpi 0.6.0-1.1.
This file is owned by root:root, with mode 0o755.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 | #!/usr/bin/perl -w
# t-avg - Calculates statistics of a **series** of piv data.
#
# Copyright (C) 2002 Gerber van der Graaf
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2, or (at your option)
# any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software Foundation,
# Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
#------------------------------------------------------------------
$VERSION = q$Id: t-avg.pl,v 1.6 2006/03/04 12:37:08 gerber Exp $;
$HELP = "Calculates time-avaraged mean and rms from a **series** of piv data";
$USAGE = "gpiv_t-avg [-h|-help] [-n|-none] [-p|-print] [-relrms] [-uv] [-v|-version]
[-dn|dir_name] [-fd|-first_dir N] [-ld|-last_dir N] [-dx|dir_prefix]
[-fb|-file_basename] [-fext|-file_extension] [-ff|-first_file N]
[-fl|-last_file N] [-fx|file_prefix]
keys:
-h: on-line help
-n: suppresses real execution
-p: prints process parameters/variables to stdout
-relrms: prints relative rms, related to the absolute mean value
-uv: writes u and v values to separate outputs
-v: prints version
-dn: directory name to be processed
-df: starting at first directory number N
-dl: ending at last directory number N
-dx: prefix numbering to directory base name
-fb: file base name to be processed
-fext: add an extension after the file basename + number (without
leading \".\")
-ff: starting at first file number N
-fl: ending at last file number N
-fx: prefix numbering to file base name
";
#----------------- Command arguments handling ----------
$opt_h = 0;
$opt_n = 0;
$opt_p = 0;
$opt_uv = 0;
$opt_relrms = 0;
$opt_v = 0;
$opt_dx = 0;
$opt_fx = 0;
use Getopt::Long;
GetOptions('h|help', 'n|none', 'p|print', 'v|version',
'relrms',
'uv',
'dx|dir_prefix', 'fx|file_prefix',
'fb|file_basename=s' => \$file_base_name,
'fext|file_extension=s' => \$file_ext,
'ff|first_file=i' => \$file_first_nr,
'fl|last_file=i' => \$file_last_nr,
'dn|dir_name=s' => \$dir_base_name,
'df|dir_first=i' => \$dir_first_nr,
'dl|dir_last=i' => \$dir_last_nr);
if ($opt_h) {
print ("$HELP\n");
print ("$USAGE\n");
exit;
}
if ($opt_v) {
print ("$VERSION\n");
exit;
}
if ($#ARGV != -1) {
print ("Usage: $USAGE\n");
exit;
}
#---------------------------------------------------- Initializing variables
$first = 1;
$k = 0;
if (!$file_base_name) {$file_base_name = "";}
if (!$file_first_nr) {$file_first_nr = 0;} # First data set to be scanned
if (!$file_last_nr) {$file_last_nr = 2; } # Last data set to be scanned
if (!$dir_base_name) {
$dir_base_name_defined = 0; $dir_base_name = "./";
} else {
$dir_base_name_defined = 1;
}
if (!$dir_first_nr) {$dir_first_nr = 0;}
if (!$dir_last_nr) {$dir_last_nr = 0;}
if ($opt_uv) {
if ($file_ext) {
$outfile_A = $dir_base_name.$file_base_name.".ta_U".".$file_ext".".piv";
$outfile_B = $dir_base_name.$file_base_name.".ta_V".".$file_ext".".piv";
} else {
$outfile_A = $dir_base_name.$file_base_name.".ta_U.piv";
$outfile_B = $dir_base_name.$file_base_name.".ta_V.piv";
}
} else {
if ($file_ext) {
$outfile_A = $dir_base_name.$file_base_name.".ta_mean".".$file_ext".".piv";
$outfile_B = $dir_base_name.$file_base_name.".ta_sdev".".$file_ext".".piv";
} else {
$outfile_A = $dir_base_name.$file_base_name.".ta_mean.piv";
$outfile_B = $dir_base_name.$file_base_name.".ta_sdev.piv";
}
}
if ($opt_n || ($file_base_name && $opt_p)) {
print ("\nfile_base_name=$file_base_name file_first_nr=$file_first_nr file_last_nr=$file_last_nr\n");
}
if ($opt_n || ($dir_base_name && $opt_p)) {
print ("\ndir_base_name=$dir_base_name dir_nr_first=$dir_first_nr dir_nr_last=$dir_last_nr\n");
}
#--------------------------------------------- Print filenames
if ($opt_p || $opt_n) {
printf ("Output: %s\n", $file_name);
if ($opt_uv) {
printf ("Data U-file: %s\n", $outfile_A);
printf ("Data V-file: %s\n", $outfile_B);
} else {
printf ("Data mean file: %s\n", $outfile_A);
printf ("Data svar file: %s\n", $outfile_B);
}
}
#---------------------------------------- loops for directory and data sets:
for ($dir_nr = $dir_first_nr; $dir_nr <= $dir_last_nr; $dir_nr++) {
if ($dir_base_name_defined) {
if (!$opt_dx) {
$target_dir = $dir_base_name.$dir_nr."/";
} else {
$target_dir = $dir_nr.$dir_base_name."/";
}
} else {
$target_dir = $dir_base_name;
}
if ($opt_p || $opt_n) {
printf ("\ntarget_dir[%d] = %s\n", $dir_nr, $target_dir);
}
for ($file_nr = $file_first_nr; $file_nr <= $file_last_nr; $file_nr++) {
if (!$opt_fx) {
#print ("append\n");
$file_name=$target_dir.$file_base_name.$file_nr;
if ($file_ext) {$file_name=$file_name.".$file_ext"};
} else {
#print ("prefix\n");
$file_name=$target_dir.$file_nr.$file_base_name;
if ($file_ext) {$file_name=$file_name.".$file_ext"};
}
$file_name_piv = $file_name.".piv";
$infile = $file_name_piv;
#--------------------------------------------- Print input filename
if ($opt_p || $opt_n) {
printf ("Data piv file: %s\n", $file_name_piv);
}
if (!$opt_n) {
open (IN,"$infile") || die 'can\'t open $infile';
#----------------------------------------- Starts reading the lines with data
while (<IN>) {
chomp;
#---------------------------------- skip blank lines or ; sign at first col
if ($_ ne "" && (!/^(\#|;)/)) {
#------------------------------------ substitutes eventually komma's by dot
#------------------------------ for example at LaVision's PIV program DaVis
s/,/./g;
#---------------------------------------- removes eventually leading spaces
s/^ *//;
#------------------- splitting line $_ up in words; #same as @words = split
@words = split(/\s+/,$_);
#------------ As the total number of data is unknown in forward, the variables
# are set to zero when the first data set is scanned. Provided all data sets
# will have the same number (and positions!!)
if ($first) {
$total_dx[$k] = 0;
$total_dy[$k] = 0;
$total_dx_2[$k] = 0;
$total_dy_2[$k] = 0;
$count[$k] = 0;
}
$x_pos[$k] = $words[0];
$y_pos[$k] = $words[1];
$dx[$k] = $words[2];
$dy[$k] = $words[3];
$snr[$k] = $words[4];
$peak_nr[$k] = $words[5];
# to avoid warning message $snr is used twice
$snr[$k] = 0.0;
#------------------------------------------------ Take sum and sum of squares
if ($peak_nr[$k] != -1) {
$total_dx[$k] += $dx[$k];
$total_dy[$k] += $dy[$k];
$total_dx_2[$k] += $dx[$k]**2;
$total_dy_2[$k] += $dy[$k]**2;
$count[$k]++;
}
$k++;
}
}
if ($first) {
$ndata = $k;
if ($opt_p ) {
printf ("\ndata = %d\n", $ndata);
}
$first = 0;
}
$k = 0;
close (IN) || die 'can\'t close $infile';
}
}
}
#-------------------------------------- Calculating mean and rms or variances
if (!$opt_n) {
for ($k=0; $k < $ndata; $k++) {
if ($count[$k] > 1) {
$mean_x[$k]= $total_dx[$k]/$count[$k];
$svar_x[$k]= sqrt(($total_dx_2[$k]/($count[$k]-1) -
$total_dx[$k]**2 / ($count[$k]*($count[$k]-1))));
$mean_y[$k]= $total_dy[$k]/$count[$k];
$svar_y[$k]= sqrt(($total_dy_2[$k]/($count[$k]-1) -
$total_dy[$k]**2 / ($count[$k]*($count[$k]-1))));
} else {
$mean_x[$k] = $k;
$svar_x[$k] = 0.;
$mean_y[$k] = $k;
$svar_y[$k] = 0.;
$count[$k] = -1;
}
$mean_xy[$k] = sqrt(($mean_x[$k]**2 + $mean_y[$k]**2));
$svar_xy[$k] = sqrt(($svar_x[$k]**2 + $svar_y[$k]**2));
}
#----------------------------------------------------------- Writing results
open (OUT_A,">$outfile_A") || die 'can\'t open $outfile_A';
open (OUT_B,">$outfile_B") || die 'can\'t open $outfile_B';
if ($opt_uv) {
printf(OUT_A "#$VERSION");
printf(OUT_A "\n# x y x_mean x_sdev count");
printf(OUT_B "#$VERSION");
printf(OUT_B "\n# x y y_mean y_sdev count");
for ($k=0; $k < $ndata; $k++) {
printf(OUT_A "\n%-f %-f %-f %-f %-f %-2d",
$x_pos[$k], $y_pos[$k], $mean_x[$k], $svar_x[$k], $count[$k]);
printf(OUT_B "\n%-f %-f %-f %-f %-f %-2d",
$x_pos[$k], $y_pos[$k], $mean_y[$k], $svar_y[$k], $count[$k]);
}
} elsif ($opt_relrms) {
printf(OUT_A "#$VERSION");
printf(OUT_A "\n# x y x_mean y_mean sdev/mean count");
printf(OUT_B "#$VERSION");
printf(OUT_B "\n# x y x_sdev y_sdev sdev sdev/mean count");
for ($k=0; $k < $ndata; $k++) {
printf(OUT_A "\n%f %f %f %f %f %d",
$x_pos[$k], $y_pos[$k], $mean_x[$k], $mean_y[$k],
# $mean_xy[$k],
$svar_xy[$k] / $mean_xy[$k],
$count[$k]);
printf(OUT_B "\n%f %f %f %f %f %f %d",
$x_pos[$k], $y_pos[$k], $svar_x[$k], $svar_y[$k],
$svar_xy[$k], $svar_xy[$k] / $mean_xy[$k], $count[$k]);
}
} else {
printf(OUT_A "#$VERSION");
printf(OUT_A "\n# x y x_mean y_mean sdev count");
printf(OUT_B "#$VERSION");
printf(OUT_B "\n# x y x_sdev y_sdev sdev count");
for ($k=0; $k < $ndata; $k++) {
printf(OUT_A "\n%-f %-f %-f %-f %-f %-2d",
$x_pos[$k], $y_pos[$k], $mean_x[$k], $mean_y[$k],
$svar_xy[$k], $count[$k]);
printf(OUT_B "\n%-f %-f %-f %-f %-f %-2d",
$x_pos[$k], $y_pos[$k], $svar_x[$k], $svar_y[$k],
$svar_xy[$k], $count[$k]);
}
}
close (OUT_A) || die 'can\'t close $outfile';
close (OUT_B) || die 'can\'t close $outfile';
}
#
# Thats all folks.
#
|