Kurt Schwehr
Copyright (C) 2003
kdschwehr _at_ ucsd dot edu
Viz Lecture Series
Lecture 4
$Revision: 531 $
CONTENTS
* DEMOS OF THE DAY
* PRINTING WITH ENSCRIPT
* DRAGGERS
* MAKING ANIMATED GIFS
* PUTTING TEXT ON IMAGES
* NETSCAPE/MOZILLA URL LOADING
* SANTA BARBARA BASIN
* INSTALLING OPEN INVENTOR ON MAC OSX 10.2
* EMACS QUICK STARTUP
* 4DWM TRICKS
* WHAT THE HECK IS THIS?
* NASA SPACECRAFT MODELS
----------------------------------------------------------------------
* DEMOS OF THE DAY
Thanks to Carrie Nishimura for this demo. It uses a blink node to
animate a bouncing ball.
ivview blinker.iv
----------------------------------------------------------------------
* PRINTING WITH ENSCRIPT
Enscript is a very handy unix printing tool that can print two pages
per side of a printer. It is only good for straight ASCII text, but I
like it and it can save a lot of trees. Here is a sample of how to
use it:
enscript -2Gr -Pmyprinter VizLecture4.txt
enscript -2Gr -pVizLecture4.ps
gv VizLecture4.ps
gs VizLecture4.ps
ghostview VizLecture4.ps
xpsview VizLecture4.ps
pageview VizLecture4.ps
For more info:
man enscript
----------------------------------------------------------------------
* DRAGGERS
Draggers are a great way to control actions inside of a 3D world. We
will first go through how to use a 1 dimensional dragger to create a
changeable sea surface. We will then move on to some of the other
draggers. First, we need a simple 1D dragger (1d-dragger.iv):
#Inventor V2.1 ascii
Translate1Dragger {}
That's all. This gives you an arrow on the screen. Press the "Esc"
key to change from the hand to the arrow. Then click and hold the
left mouse button on the arrow. Now drag it around. Now we can add
some text that will give us feedback as to what the dragger is putting
out (1d-drag-text.iv):
DEF dragger Translate1Dragger {}
Text3 { string "Offset" =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oa = C[0]"
}
. oa
}
This above sample is pretty tricky as it introduces a bunch of
concepts altogether. The first is the DEF keyword. This allows us to
reuse an object later on in an OpenInventor file. It can also be used
to save memory if an object is used many times. Here is a simple def
example that reuses one sphere node so you will see more than one
sphere (def.iv):
DEF s Sphere {}
Translation { translation 0 0 3 }
USE s
The second big concept is the calculator node. Calculators are very
handy. They can be used to do all kinds of math within the scene
graph. Here we just use a calculator to get the X axis of the
draggers translation. Calculators use a set of registers for
calculations. Here we use the "oa" register to store the X axis and
then send it back to set the string's value with the X offset.
If we want to translate the text with the slider we need to add a
translation that is controlled by another calculator. This looks
pretty much like the one we used for the text, but instead of a single
float value we need a 3D vector, so we need to use the oA register and
the vec3f function to fill it (1d-drag-object.iv):
Transform {
translation 0 0 0 =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oA = vec3f(C[0],0,0)"
}
. oA
}
To show an example of this all in action, take a look at
ocean-slider.iv which shows a sea level rise and fall in the Z
direction.
Another task you may want to do is to scale an object with a dragger.
Turns out that it is almost the same as what we just did above.
Instead of a translation, we will use a scaleFactor that is controlled
by the draggers translation (scaleObject):
Transform {
scaleFactor 1 1 1 =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oA = vec3f(C[0],C[0],C[0])"
}
. oA
}
For more information, see:
man SoTranslate1Dragger
man SoCalculator
man SoTransform
----------------------------------------------------------------------
* MAKING ANIMATED GIFS
A common task that people like to do is to produce animated gifs.
There are lots of expensive packages that can do this, but let's look
at a simple one called "gifsicle":
http://www.lcdf.org/gifsicle/
We will make a simple animation of the NOAA images of the east coast
blackout of Aug 2003. First, grab the images from NOAA:
http://www.noaanews.noaa.gov/stories/s2015.htm
wget http://www.noaanews.noaa.gov/nightlights/blackout081403-20hrsbefore-text.jpg
wget http://www.noaanews.noaa.gov/nightlights/blackout081503-7hrsafter-text.jpg
Convert the images to gif from jpg:
jpegtopnm blackout081403-20hrsbefore-text.jpg | ppmquant 255 | ppmtogif > blackout081403-20hrsbefore-text.gif
jpegtopnm blackout081503-7hrsafter-text.jpg | ppmquant 255 | ppmtogif > blackout081503-7hrsafter-text.gif
Now make the animation:
gifsicle --colors 256 --delay=100 --loop blackout081*.gif > blackout-anim.gif
There are many more options to gifsicle. Check the web or man page:
setenv MANPATH ${MANPATH}:~schwehr/IRIX64-6.5/man
man gifsicle
----------------------------------------------------------------------
* PUTTING TEXT ON IMAGES
Sometimes you need to be able to put some text on an image. The large
image processing package known as ImageMagick has just the tool for
this (along with many many other tasks). Here is a sample command
line to add text to an image:
convert -pointsize 20 -draw 'text 10,50 "Hello World"' foo.jpg foo2.jpg
display foo2.jpg
That's all there is to it. Take a look at the IBM tutorial below for
more ideas on what can be done.
For more info see:
http://www-106.ibm.com/developerworks/linux/library/l-graf/?ca=dgr-lnxw04ImageMagick
http://freshmeat.net/projects/imagemagick
----------------------------------------------------------------------
* NETSCAPE/MOZILLA URL LOADING
In the last lecture, we saw how to use the WWWAnchor node with
xcore-ivview to tell netscape how to open a URL/web page. The method
used to do this is usable from any script or program and is also
available with mozilla. So how was it done? First you must have the
browser already running. Here we will use netscape, but mozilla works
exactly the same.
netscape &
Once it is running, just run this on the command line:
netscape -remote "openURL(http://earthquake.usgs.gov/recenteqsUS/)"
Netscape should now be at the USGS US earthquake page.
For more info see:
http://wp.netscape.com/newsref/std/x-remote.html
http://www.mozilla.org/unix/remote.html
----------------------------------------------------------------------
* SANTA BARBARA BASIN
In this section, I will walk you through most of a project based on
multibeam data of the Santa Barbara Basin. Many people at SIO work in
this area, so it's probably a good one to work on. I will not be
starting with the raw mulitbeam data from a ship. Instead, we will
start with the MBARI Santa Barbara Basin CD. To produce this CD,
someone had to do a fair bit of work to clean and process the data
running such programs as mbedit to remove bad beams from each ping (a
very time consuming job!) For a tutorial on processing the raw data,
you can talk to Uta and get a copy of her tutorial for the SIO ships.
So let's jump into the processed data. Take a look at the MBARI
website:
http://www.mbari.org/data/mapping/SBBasin/default.htm
http://www.mbari.org/data/mapping/SBBasin/basin.htm
These pages show the EM300 multibeam data and some surrounding land
data. You can order the CD from MBARI or copy it from the SIO GDC.
Once you have the CD, you will find the data on Disk1 in grids/geo in
files with a ".asc" extension. These are Arc/Info ASCII format data
files. Take a look at one to see what is inside:
less santabarbA_bath.asc
head -6 santabarbA_bath.asc
ncols 3479
nrows 1371
xllcorner -121.06507
yllcorner 34.08928
cellsize 0.0004349559829
nodata_value -99999
Each file is 6 lines of header, followed by the z-value (depth) of
each point in the data set. We will use mbm_arc2grd from mbsystem and
gmt to bring in the data. mb-system can be found here:
http://www.ldeo.columbia.edu/MB-System/html5/mbsystem.html#index
ftp://ftp.ldeo.columbia.edu/pub/mbsystem
WARNING: The mbm_arc2grd perl script is only in mbsystem version 5.0
Beta or newer. Also, prior to version 5.0beta32, there is a bug in
the script. You will need to edit the script and change "-H6" to
"-H0" on the line with xyz2grd.
First there is an error in land_topo_geo. The "nodata_value" field
should be all lower case. The way it is now will cause mbm_arc2grd to
not work
head -6 land_topo_geo.asc
ncols 2111
nrows 1120
xllcorner -120.845477
yllcorner 33.858087
cellsize 0.00099396289027745
NODATA_value -9999
Create a small perl script to fix the file (fixAsc.pl):
#!/usr/bin/perl -pi
s/^NODATA/nodata/;
Make the script executable and use it:
chmod +x fixAsc.pl
./fixAsc land_topo_geo.asc
head -6 land_topo_geo.asc
Now we are all set to create grid files out of each ascii Arc/Info
file. On the IGPP fishnet, you will need to be logged in to a sun
that mounts the GDC's tools to use mbm_arc2grd (note the export line
in the script) and create the arc2grdAll.bash script:
#!/bin/bash
export PATH=/net/gdcraid/gdcportal/bin/mbsystem:${PATH}
for file in *.asc
do
echo Converting $file
mbm_arc2grd -I$file -O${file%%.asc}.grd
done
Now run the script. This will take a bit of time and uses an extra
200 MB of disk space.
chmod +x arc2grdAll.bash
./arc2grdAll.bash
You now have a grd file for each original ascii file:
grdinfo land_topo_geo.grd
Since these files are pretty big, we can use our OpenInventor skills
to write a little script that will produce a labeled bounding box for
each grd file (bbox.bash):
#!/bin/bash
for file in *bath.grd *geo.grd
do
echo processing $file
X=`grdinfo ${file}|grep x_min`
Y=`grdinfo ${file}|grep y_min`
XMIN=`echo $X | awk '{print $3}'`
XMAX=`echo $X | awk '{print $5}'`
YMIN=`echo $Y | awk '{print $3}'`
YMAX=`echo $Y | awk '{print $5}'`
XAVE=`echo $XMIN $XMAX | awk '{print ($1+$2)/2.}'`
YAVE=`echo $YMIN $YMAX | awk '{print ($1+$2)/2.}'`
out=${file%%.grd}-bbox.iv
cat < $out
#Inventor V2.1 ascii
Separator {
Coordinate3 { point [
$XMIN $YMIN 0.,
$XMAX $YMIN 0.,
$XMAX $YMAX 0.,
$XMIN $YMAX 0.
] }
BaseColor { rgb [ 1 0 0] }
IndexedLineSet { coordIndex [ 0, 1, 2, 3, 0 ] }
Translation { translation $XAVE $YAVE 0.05 }
Scale { scaleFactor .002 .002 .002 }
BaseColor { rgb [ 1 1 1] }
Text3 { string "${file%%.grd}" }
}
EOF
done
I think this is a pretty clever little script. It draws a square
around each grd region and labels the file name. It produces
something like this (land_topo_geo-bbox.iv):
#Inventor V2.1 ascii
Separator {
Coordinate3 { point [
-120.845 33.8586 0.,
-118.747 33.8586 0.,
-118.747 34.9708 0.,
-120.845 34.9708 0.
] }
BaseColor { rgb [ 1 0 0] }
IndexedLineSet { coordIndex [ 0, 1, 2, 3, 0 ] }
Translation { translation -119.796 34.4147 0.05 }
Scale { scaleFactor .002 .002 .002 }
BaseColor { rgb [ 1 1 1] }
Text3 { string "land_topo_geo" }
}
So run the script and look at the output:
chmod +x bbox.bash
./bbox.bash
ivview *-bbox.iv
Now that we know where everything is, we can create some smaller grd
files and 3D terrains. The default grds are really large, so we will
cut them down by a factor of 4. Here is a script to do this for all
of the files (makeTerrains.bash):
#!/bin/bash
gmtset D_FORMAT = %l.10g # Fix precision problem with grd2xyz
for file in *geo.grd *bath.grd
do
echo processing $file
X=`grdinfo ${file}|grep x_min`
Y=`grdinfo ${file}|grep y_min`
XINC=`echo $X | awk '{print $7}'`
YINC=`echo $Y | awk '{print $7}'`
grdsample $file -G${file%%.grd}_a4.grd \
-I`echo $XINC | awk '{print $1*4}'`
grd2xyz ${file%%.grd}_a4.grd > ${file%%.grd}_a4.xyz
terrain ${file%%.grd}_a4.xyz ${file%%.grd}_a4.iv
done
Now take a look at one of the files:
ivview land_topo_geo_a4.iv
Here is a handy way to view all of the files (viewAll.bash):
#!/bin/bash
for file in *_a4.iv
do
b=${file%%_a4.iv}
ivview -name "$b" $file $b-bbox.iv
done
For my thesis, I'm focusing on the Gaviota Slide. I want the highest
resolution model just in the the area of the slide. It is just to the
west of the much larger Golita slide. The area of interest is:
120 deg 9 min West -- 120 deg 3 min West
34 deg 20 min North -- 34 deg 24 min North
Which is a GMT region of (see "man grdimage"):
-R-120:09/-120:03/34:20/34:24
or
-R-120.15/-120.05/34.33333333/34.4
Looking thought the IV models of the terrains, the Gaviota Slide is
in the following grd files:
santabarbA_bath - Lower resolution wide area model
santabarbC_bath - The S.B. Basin
santabarbF_bath - Northern wall of the basin... highest resolution
Based on this, we will start with the "F" file and cut the grid down
to right around the slide. Now use gridcut to get the piece, but we
have to do a little work to handle the cell size of the grd file format.
If we just use our new region straight off, grdcut will fail:
grdcut santabarbF_bath.grd -Ggaviota.grd -R-120.15/-120.05/34.33333333/34.4
grdcut: GMT ERROR: (x_max-x_min) must equal (NX + eps) * x_inc), where NX is an integer and |eps| <= 0.0001.
grdcut: GMT ERROR: (y_max-y_min) must equal (NY + eps) * y_inc), where NY is an integer and |eps| <= 0.0001.
We are lucky here that the grid file has the same spacing in the X and
Y dimensions. You can view this as the x_inc and y_inc fields with
grdinfo
grdinfo santabarbF_bath.grd | grep inc | grep inc | awk '{print $6, $7}'
x_inc: 0.000108738
y_inc: 0.000108738
Trying to compute these values is not very fun, but luckily mb_system
(version 5 and newer) provides a tool to do this for us. mbm_grdcut
will try to give a best fit region. The name of the program may be
mbm_grdcut.pl, mbm_grdcut, or mbm_grdcut.perl depending on your system
configuration.
mbm_grdcut.pl -IsantabarbF_bath.grd -Ogaviota.grd -R-120:09/-120:03/34:20/34:24 -V
diffx: 1454.00932392159 1454
Program mbm_grdcut status:
Input GRD file: santabarbF_bath.grd
Output GRD file: gaviota.grd
Input Grid bounds: -120.3081056 -119.7626774 34.30966037 34.4629805
Input grid dimensions: 5017 1411
Grid cell sizes: 0.0001087376796 0.0001087376796
Requested Grid bounds: -120.15 -120.05 34.3333333333333 34.4
Output Grid bounds: -120.150001013862 -120.04996234863 34.3332564464732 34.4000213817476
Output grid dimensions: 921 615
Running grdcut...
grdcut santabarbF_bath.grd -Ggaviota.grd -R-120.150001013862/-120.04996234863/34.3332564464732/34.4000213817476 -V
grdcut: File spec: W E S N dx dy nx ny:
grdcut: Old: -120.308 -119.763 34.3097 34.463 0.000108738 0.000108738 5017 1411
grdcut: New: -120.15 -120.05 34.3333 34.4 0.000108738 0.000108738 921 615
We can now build a 3D model of the Gaviota slide:
grd2xyz gaviota.grd > gaviota.xyz
terrain gaviota.xyz gaviota.iv
ivview -name "Gaviota Slide" gaviota.iv
NOTE: There is a bug with terrain generation of very high resolution
grd files. This will appear as stripes in the models. This problem
exists as of Aug 2003.
How about marking the locations of the cores collected in the area?
The USGS has a database of cruise metadata called InfoBank. The page
for the cruise that I am working with is here and the core locations
are under "Samples":
http://walrus.wr.usgs.gov/infobank/v/v181sc/html/v-1-81-sc.meta.html
We can grab all the core records like this
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.8g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.9g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.10g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.11g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.12g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.13g2
A small script will create a table of the core locations (getCoreLoc.bash):
#!/bin/bash
for file in v-1-81-sc.core.gravitycore.*g2
do
lat=`grep Latitude $file | awk '{print $3}'`
lon=`grep Longitude $file | awk '{print $3}'`
id=`grep 'Sample ID' $file | awk '{print $4}'`
depth=-`grep 'Water Depth' $file | awk '{print $4}'`
length=`grep 'Core Length' $file | awk '{print $4}'`
echo $lat $lon $depth ${id}_${length}cm >> v-1-81-sc.corelocations
done
From these, we build up a table of locations. However, the depth for
09G2 or 08G2 looks suspicious!
34.36500 -120.10700 -477 "10G2_177cm"
34.36700 -120.10800 -455 "11G2_155cm"
34.38200 -120.10800 -336 "12G2_186cm"
34.37300 -120.11800 -396 "13G2_178cm"
34.34700 -120.10700 -545 "8G2_148cm"
34.36000 -120.10300 -977 "9G2_149cm"
Using this table we can create labels that mark the location of cores:
cat v-1-81-sc.corelocations | awk '{print $1,$2,$4}' > gaviota-coreloc.txt
xytext2iv gaviota-coreloc.txt gaviota-coreloc
From there, you will need to edit the scale and translations to make
the xytext2iv program work at this scale. It was meant for California
sized objects.
----------------------------------------------------------------------
* INSTALLING OPEN INVENTOR ON MAC OSX 10.2
Thanks to Davis Thomsen for the initial document after we attacked his new
laptop.
Installing Inventor on a Ti G4 Powerbook with OSX
1) Install the developer tools that come with OSX. This can be done a
couple of ways:
a) Install from the disk that comes with OSX when bought at the store
b) Install from the packages already on the computer. This way
involves installing these packages in a specific order. In
/Applications/Installers/Developer Tools/Packages
i. install BSDSDK.pkg
ii. install DevTools.pkg
iii. finally install DevSDK.pkg
c) It is possible to download these from the apple website.
* http://connect.apple.com (can get a free login)
* Click Download Software
* Click "Developer Tools"
* December 2002 Mac OS X Developer Tools
Click "Download" for Dec 2002 Mac OS X Developer Tools
WARNING: This is over 300 MB!
2) Install X11. This can be downloaded from the apple website:
http://www.apple.com/macosx/x11/
Click Download X11 Public Beta 3
3) Install Fink and Fink Commander
Fink: http://fink.sourceforge.net/download/index.php
Fink Commander: http://finkcommander.sourceforge.net/
4) Install the X11 Developer Toolkit
At the X11 download page, http://www.apple.com/macosx/x11/download/
Near the bottom right: X11 for Mac OSX Public Beta SDK
Download SDK (4.0 Mb)
5) Use Fink Commander to download and install "system-Xfree86"
6) Use Fink Commander to download "motif" (when you type this into the
name search field, there will be three results. Install all three.
Add the library from the viz center homepage to the correct
location.
7) Use Fink Commander to download and install GMT (Generic Mapping Tool)
8) Get the Open Inventor from
http://schwehr.org/software/Darwin/OpenInventor-Works/
Untar this file and put in your HOME directory (i.e. /users/you)
Also, get the motif library on this web page and put it in
/sw/lib. This is because open inventor was build when version
2 of OpenMotif was standard in fink.
9) Add the following lines to your .cshrc:
setenv DYLD_LIBRARY_PATH ${HOME}/inventor/lib
setenv PATH ${PATH}:${HOME}/inventor/apps/tools/ivview
10) Move the "inventor" folder to ~/
11) Hopefully the software will work.
If you want to install mb_system, here are the Makefile settings that
might work for you (provided by Jasper Konter). OS X support for
MB-System should be increasing now that Dave Caress has a Titanium G4
laptop as his primary development platform.
ftp://ftp.ldeo.columbia.edu/pub/mbsystem
# **** EDIT THE PARAMETERS HERE ****
# Set the environment parameters:
$MBSYSTEM_HOME = "/sw/share/mbsystem";
$OS = "DARWIN";
$BYTESWAPPED = "NO";
$CC = "gcc";
$GRAPHICAL = "YES";
$MOTIFINCDIR = "/sw/include/Xm";
$MOTIFLIBS = "-L/sw/lib -L/usr/X11R6/lib -L/usr/X11R6/include -lXm -lXt -lX11";
$GMTVERSION = "3.4";
$GMTLIBDIR = "/sw/lib";
$GMTINCDIR = "/sw/include";
$NETCDFLIBDIR = "/sw/lib";
$NETCDFINCDIR = "/sw/include";
$LEVITUS = "/sw/share/levitus/annual";
$PROJECTIONS = "$MBSYSTEM_HOME/share/Projections.dat";
$CFLAGS = "-g -I/usr/include/X11R6 -I/usr/X11R6/include -I/sw/include -no-cpp-precomp";
$LFLAGS = "-Wl -force_flat_namespace -lm";
#
# **** LEAVE EVERYTHING ELSE BELOW ALONE ****
----------------------------------------------------------------------
* EMACS QUICK STARTUP
Emacs is a very powerful editor and I highly recommend it for all
sorts of editing and programming tasks. However, it is pretty hard to
use till you get the hang of it. Here is a quick tutorial on how to
use emacs.
The first thing to do is to setup emacs with a special configuration
file if you do not already have one. I have a sample setup that you
can use:
cd
cp ~schwehr/.emacs .
cp -r ~schwehr/.emacs-lisp .
Or here is where you can get the OpenInventor mode files yourself:
cd
mkdir .emacs-lisp
cd .emacs-lisp
wget ftp://ftp.sgi.com/sgi/inventor/emacs/iv.el
wget ftp://ftp.sgi.com/sgi/inventor/emacs/iv-mode.el
Here is the alias that I use to startup emacs in bash:
alias e='emacs -fn 6x13 -bg black -fg white'
Okay, now we are ready to see a list of commands. The first thing is
a quick note on the "meta" and "control" keys. The meta key is
written as "M-" and can be done by either pressing the Esc key once
followed by the key or by holding down "Alt" or the diamond key on
suns and pressing the key at the same time. Experiment to find out
what works on your computer. The control key is written as C- and
works just as you would expect. You can see a lot of the key
combinations if the menus are working by just by clicking on the
menu. The key equivalent is listed.
C-x C-c Exit emacs
C-x C-s Save
C-x C-w Save As a different name
C-x C-f Open a file ( hint: open "." for a directory)
C-x k Close a buffer (a.k.a. file)
C-x i Insert a file
C-g QUIT!!! Use this if stuck.
C-x 2 Split screen editing: top-bottom
C-x 3 Split screen editing: left-right
C-x 1 Single file editing
C-x b Jump to a different open file/buffer
C-s Search for string (type C-s again to find next)
C-r Search backwards
C-k Cut (Kut) to the end of the line
C-y Yank the previous cut. (M-y right after to go
back to previous cuts)
C-_ UNDO! (C-z again to keep undoing)
C-space Mark the beginning of a region
C-w Cut the region
M-w Copy the region
M-x query-replace Go through and replace one string with
another, but ask first on each one.
M-x replace-string Replace all occurrences of a string
M-x goto-line Jump to a line number
C-x C-g Jump to a line if you use my .emacs
M-x compile Compile or run a program
Compile if you use my .emacs
C-x ` Jump to next compiler error (with gcc or g77)
C-x v v RCS/CVS next (check in). Use C-c C-c when done
with entering comment
M-x info Emacs online documentation... hard to navigate!
M-x doctor Talk to a shrink.
----------------------------------------------------------------------
* 4DWM TRICKS
The default window manager on the SGI is 4Dwm. It has some keyboard
shortcuts that can be very handy with graphics applications. Give
them a try:
Alt-F3 lower
Alt-F5 restore (what does this mean?)
Alt-F7 move
Alt-F8 resize
Alt-F9 minimize
Alt-F10 maximize
I use move and lower all the time.
For more info see:
less /usr/lib/X11/system.4Dwmrc
man 4Dwm
----------------------------------------------------------------------
* WHAT THE HECK IS THIS? (From ACD)
http://www.sunpoint.net/~epa99/104_0482.JPG
----------------------------------------------------------------------
* NASA SPACECRAFT MODELS
http://samadhi.jpl.nasa.gov/models/
Here are some simple spacecraft models that you may want to take a
look at. Only two are in Inventor format. The others may be
converted, but I have not tried them. Converters on the SGI can be
found in:
ls /usr/sbin/*To*
----------------------------------------------------------------------
* PRINTING WITH ENSCRIPT
Enscript is a very handy unix printing tool that can print two pages
per side of a printer. It is only good for straight ASCII text, but I
like it and it can save a lot of trees. Here is a sample of how to
use it:
enscript -2Gr -Pmyprinter VizLecture4.txt
enscript -2Gr -pVizLecture4.ps
gv VizLecture4.ps
gs VizLecture4.ps
ghostview VizLecture4.ps
xpsview VizLecture4.ps
pageview VizLecture4.ps
For more info:
man enscript
----------------------------------------------------------------------
* DRAGGERS
Draggers are a great way to control actions inside of a 3D world. We
will first go through how to use a 1 dimensional dragger to create a
changeable sea surface. We will then move on to some of the other
draggers. First, we need a simple 1D dragger (1d-dragger.iv):
#Inventor V2.1 ascii
Translate1Dragger {}
That's all. This gives you an arrow on the screen. Press the "Esc"
key to change from the hand to the arrow. Then click and hold the
left mouse button on the arrow. Now drag it around. Now we can add
some text that will give us feedback as to what the dragger is putting
out (1d-drag-text.iv):
DEF dragger Translate1Dragger {}
Text3 { string "Offset" =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oa = C[0]"
}
. oa
}
This above sample is pretty tricky as it introduces a bunch of
concepts altogether. The first is the DEF keyword. This allows us to
reuse an object later on in an OpenInventor file. It can also be used
to save memory if an object is used many times. Here is a simple def
example that reuses one sphere node so you will see more than one
sphere (def.iv):
DEF s Sphere {}
Translation { translation 0 0 3 }
USE s
The second big concept is the calculator node. Calculators are very
handy. They can be used to do all kinds of math within the scene
graph. Here we just use a calculator to get the X axis of the
draggers translation. Calculators use a set of registers for
calculations. Here we use the "oa" register to store the X axis and
then send it back to set the string's value with the X offset.
If we want to translate the text with the slider we need to add a
translation that is controlled by another calculator. This looks
pretty much like the one we used for the text, but instead of a single
float value we need a 3D vector, so we need to use the oA register and
the vec3f function to fill it (1d-drag-object.iv):
Transform {
translation 0 0 0 =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oA = vec3f(C[0],0,0)"
}
. oA
}
To show an example of this all in action, take a look at
ocean-slider.iv which shows a sea level rise and fall in the Z
direction.
Another task you may want to do is to scale an object with a dragger.
Turns out that it is almost the same as what we just did above.
Instead of a translation, we will use a scaleFactor that is controlled
by the draggers translation (scaleObject):
Transform {
scaleFactor 1 1 1 =
Calculator {
C 0 0 0 =
USE dragger
. translation
expression "oA = vec3f(C[0],C[0],C[0])"
}
. oA
}
For more information, see:
man SoTranslate1Dragger
man SoCalculator
man SoTransform
----------------------------------------------------------------------
* MAKING ANIMATED GIFS
A common task that people like to do is to produce animated gifs.
There are lots of expensive packages that can do this, but let's look
at a simple one called "gifsicle":
http://www.lcdf.org/gifsicle/
We will make a simple animation of the NOAA images of the east coast
blackout of Aug 2003. First, grab the images from NOAA:
http://www.noaanews.noaa.gov/stories/s2015.htm
wget http://www.noaanews.noaa.gov/nightlights/blackout081403-20hrsbefore-text.jpg
wget http://www.noaanews.noaa.gov/nightlights/blackout081503-7hrsafter-text.jpg
Convert the images to gif from jpg:
jpegtopnm blackout081403-20hrsbefore-text.jpg | ppmquant 255 | ppmtogif > blackout081403-20hrsbefore-text.gif
jpegtopnm blackout081503-7hrsafter-text.jpg | ppmquant 255 | ppmtogif > blackout081503-7hrsafter-text.gif
Now make the animation:
gifsicle --colors 256 --delay=100 --loop blackout081*.gif > blackout-anim.gif
There are many more options to gifsicle. Check the web or man page:
setenv MANPATH ${MANPATH}:~schwehr/IRIX64-6.5/man
man gifsicle
----------------------------------------------------------------------
* PUTTING TEXT ON IMAGES
Sometimes you need to be able to put some text on an image. The large
image processing package known as ImageMagick has just the tool for
this (along with many many other tasks). Here is a sample command
line to add text to an image:
convert -pointsize 20 -draw 'text 10,50 "Hello World"' foo.jpg foo2.jpg
display foo2.jpg
That's all there is to it. Take a look at the IBM tutorial below for
more ideas on what can be done.
For more info see:
http://www-106.ibm.com/developerworks/linux/library/l-graf/?ca=dgr-lnxw04ImageMagick
http://freshmeat.net/projects/imagemagick
----------------------------------------------------------------------
* NETSCAPE/MOZILLA URL LOADING
In the last lecture, we saw how to use the WWWAnchor node with
xcore-ivview to tell netscape how to open a URL/web page. The method
used to do this is usable from any script or program and is also
available with mozilla. So how was it done? First you must have the
browser already running. Here we will use netscape, but mozilla works
exactly the same.
netscape &
Once it is running, just run this on the command line:
netscape -remote "openURL(http://earthquake.usgs.gov/recenteqsUS/)"
Netscape should now be at the USGS US earthquake page.
For more info see:
http://wp.netscape.com/newsref/std/x-remote.html
http://www.mozilla.org/unix/remote.html
----------------------------------------------------------------------
* SANTA BARBARA BASIN
In this section, I will walk you through most of a project based on
multibeam data of the Santa Barbara Basin. Many people at SIO work in
this area, so it's probably a good one to work on. I will not be
starting with the raw mulitbeam data from a ship. Instead, we will
start with the MBARI Santa Barbara Basin CD. To produce this CD,
someone had to do a fair bit of work to clean and process the data
running such programs as mbedit to remove bad beams from each ping (a
very time consuming job!) For a tutorial on processing the raw data,
you can talk to Uta and get a copy of her tutorial for the SIO ships.
So let's jump into the processed data. Take a look at the MBARI
website:
http://www.mbari.org/data/mapping/SBBasin/default.htm
http://www.mbari.org/data/mapping/SBBasin/basin.htm
These pages show the EM300 multibeam data and some surrounding land
data. You can order the CD from MBARI or copy it from the SIO GDC.
Once you have the CD, you will find the data on Disk1 in grids/geo in
files with a ".asc" extension. These are Arc/Info ASCII format data
files. Take a look at one to see what is inside:
less santabarbA_bath.asc
head -6 santabarbA_bath.asc
ncols 3479
nrows 1371
xllcorner -121.06507
yllcorner 34.08928
cellsize 0.0004349559829
nodata_value -99999
Each file is 6 lines of header, followed by the z-value (depth) of
each point in the data set. We will use mbm_arc2grd from mbsystem and
gmt to bring in the data. mb-system can be found here:
http://www.ldeo.columbia.edu/MB-System/html5/mbsystem.html#index
ftp://ftp.ldeo.columbia.edu/pub/mbsystem
WARNING: The mbm_arc2grd perl script is only in mbsystem version 5.0
Beta or newer. Also, prior to version 5.0beta32, there is a bug in
the script. You will need to edit the script and change "-H6" to
"-H0" on the line with xyz2grd.
First there is an error in land_topo_geo. The "nodata_value" field
should be all lower case. The way it is now will cause mbm_arc2grd to
not work
head -6 land_topo_geo.asc
ncols 2111
nrows 1120
xllcorner -120.845477
yllcorner 33.858087
cellsize 0.00099396289027745
NODATA_value -9999
Create a small perl script to fix the file (fixAsc.pl):
#!/usr/bin/perl -pi
s/^NODATA/nodata/;
Make the script executable and use it:
chmod +x fixAsc.pl
./fixAsc land_topo_geo.asc
head -6 land_topo_geo.asc
Now we are all set to create grid files out of each ascii Arc/Info
file. On the IGPP fishnet, you will need to be logged in to a sun
that mounts the GDC's tools to use mbm_arc2grd (note the export line
in the script) and create the arc2grdAll.bash script:
#!/bin/bash
export PATH=/net/gdcraid/gdcportal/bin/mbsystem:${PATH}
for file in *.asc
do
echo Converting $file
mbm_arc2grd -I$file -O${file%%.asc}.grd
done
Now run the script. This will take a bit of time and uses an extra
200 MB of disk space.
chmod +x arc2grdAll.bash
./arc2grdAll.bash
You now have a grd file for each original ascii file:
grdinfo land_topo_geo.grd
Since these files are pretty big, we can use our OpenInventor skills
to write a little script that will produce a labeled bounding box for
each grd file (bbox.bash):
#!/bin/bash
for file in *bath.grd *geo.grd
do
echo processing $file
X=`grdinfo ${file}|grep x_min`
Y=`grdinfo ${file}|grep y_min`
XMIN=`echo $X | awk '{print $3}'`
XMAX=`echo $X | awk '{print $5}'`
YMIN=`echo $Y | awk '{print $3}'`
YMAX=`echo $Y | awk '{print $5}'`
XAVE=`echo $XMIN $XMAX | awk '{print ($1+$2)/2.}'`
YAVE=`echo $YMIN $YMAX | awk '{print ($1+$2)/2.}'`
out=${file%%.grd}-bbox.iv
cat <
Now that we know where everything is, we can create some smaller grd
files and 3D terrains. The default grds are really large, so we will
cut them down by a factor of 4. Here is a script to do this for all
of the files (makeTerrains.bash):
#!/bin/bash
gmtset D_FORMAT = %l.10g # Fix precision problem with grd2xyz
for file in *geo.grd *bath.grd
do
echo processing $file
X=`grdinfo ${file}|grep x_min`
Y=`grdinfo ${file}|grep y_min`
XINC=`echo $X | awk '{print $7}'`
YINC=`echo $Y | awk '{print $7}'`
grdsample $file -G${file%%.grd}_a4.grd \
-I`echo $XINC | awk '{print $1*4}'`
grd2xyz ${file%%.grd}_a4.grd > ${file%%.grd}_a4.xyz
terrain ${file%%.grd}_a4.xyz ${file%%.grd}_a4.iv
done
Now take a look at one of the files:
ivview land_topo_geo_a4.iv
Here is a handy way to view all of the files (viewAll.bash):
#!/bin/bash
for file in *_a4.iv
do
b=${file%%_a4.iv}
ivview -name "$b" $file $b-bbox.iv
done
For my thesis, I'm focusing on the Gaviota Slide. I want the highest
resolution model just in the the area of the slide. It is just to the
west of the much larger Golita slide. The area of interest is:
120 deg 9 min West -- 120 deg 3 min West
34 deg 20 min North -- 34 deg 24 min North
Which is a GMT region of (see "man grdimage"):
-R-120:09/-120:03/34:20/34:24
or
-R-120.15/-120.05/34.33333333/34.4
Looking thought the IV models of the terrains, the Gaviota Slide is
in the following grd files:
santabarbA_bath - Lower resolution wide area model
santabarbC_bath - The S.B. Basin
santabarbF_bath - Northern wall of the basin... highest resolution
Based on this, we will start with the "F" file and cut the grid down
to right around the slide. Now use gridcut to get the piece, but we
have to do a little work to handle the cell size of the grd file format.
If we just use our new region straight off, grdcut will fail:
grdcut santabarbF_bath.grd -Ggaviota.grd -R-120.15/-120.05/34.33333333/34.4
grdcut: GMT ERROR: (x_max-x_min) must equal (NX + eps) * x_inc), where NX is an integer and |eps| <= 0.0001.
grdcut: GMT ERROR: (y_max-y_min) must equal (NY + eps) * y_inc), where NY is an integer and |eps| <= 0.0001.
We are lucky here that the grid file has the same spacing in the X and
Y dimensions. You can view this as the x_inc and y_inc fields with
grdinfo
grdinfo santabarbF_bath.grd | grep inc | grep inc | awk '{print $6, $7}'
x_inc: 0.000108738
y_inc: 0.000108738
Trying to compute these values is not very fun, but luckily mb_system
(version 5 and newer) provides a tool to do this for us. mbm_grdcut
will try to give a best fit region. The name of the program may be
mbm_grdcut.pl, mbm_grdcut, or mbm_grdcut.perl depending on your system
configuration.
mbm_grdcut.pl -IsantabarbF_bath.grd -Ogaviota.grd -R-120:09/-120:03/34:20/34:24 -V
diffx: 1454.00932392159 1454
Program mbm_grdcut status:
Input GRD file: santabarbF_bath.grd
Output GRD file: gaviota.grd
Input Grid bounds: -120.3081056 -119.7626774 34.30966037 34.4629805
Input grid dimensions: 5017 1411
Grid cell sizes: 0.0001087376796 0.0001087376796
Requested Grid bounds: -120.15 -120.05 34.3333333333333 34.4
Output Grid bounds: -120.150001013862 -120.04996234863 34.3332564464732 34.4000213817476
Output grid dimensions: 921 615
Running grdcut...
grdcut santabarbF_bath.grd -Ggaviota.grd -R-120.150001013862/-120.04996234863/34.3332564464732/34.4000213817476 -V
grdcut: File spec: W E S N dx dy nx ny:
grdcut: Old: -120.308 -119.763 34.3097 34.463 0.000108738 0.000108738 5017 1411
grdcut: New: -120.15 -120.05 34.3333 34.4 0.000108738 0.000108738 921 615
We can now build a 3D model of the Gaviota slide:
grd2xyz gaviota.grd > gaviota.xyz
terrain gaviota.xyz gaviota.iv
ivview -name "Gaviota Slide" gaviota.iv
NOTE: There is a bug with terrain generation of very high resolution
grd files. This will appear as stripes in the models. This problem
exists as of Aug 2003.
How about marking the locations of the cores collected in the area?
The USGS has a database of cruise metadata called InfoBank. The page
for the cruise that I am working with is here and the core locations
are under "Samples":
http://walrus.wr.usgs.gov/infobank/v/v181sc/html/v-1-81-sc.meta.html
We can grab all the core records like this
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.8g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.9g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.10g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.11g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.12g2
wget http://walrus.wr.usgs.gov/infobank/v/v181sc/samp/v-1-81-sc.core.gravitycore.13g2
A small script will create a table of the core locations (getCoreLoc.bash):
#!/bin/bash
for file in v-1-81-sc.core.gravitycore.*g2
do
lat=`grep Latitude $file | awk '{print $3}'`
lon=`grep Longitude $file | awk '{print $3}'`
id=`grep 'Sample ID' $file | awk '{print $4}'`
depth=-`grep 'Water Depth' $file | awk '{print $4}'`
length=`grep 'Core Length' $file | awk '{print $4}'`
echo $lat $lon $depth ${id}_${length}cm >> v-1-81-sc.corelocations
done
From these, we build up a table of locations. However, the depth for
09G2 or 08G2 looks suspicious!
34.36500 -120.10700 -477 "10G2_177cm"
34.36700 -120.10800 -455 "11G2_155cm"
34.38200 -120.10800 -336 "12G2_186cm"
34.37300 -120.11800 -396 "13G2_178cm"
34.34700 -120.10700 -545 "8G2_148cm"
34.36000 -120.10300 -977 "9G2_149cm"
Using this table we can create labels that mark the location of cores:
cat v-1-81-sc.corelocations | awk '{print $1,$2,$4}' > gaviota-coreloc.txt
xytext2iv gaviota-coreloc.txt gaviota-coreloc
From there, you will need to edit the scale and translations to make
the xytext2iv program work at this scale. It was meant for California
sized objects.
----------------------------------------------------------------------
* INSTALLING OPEN INVENTOR ON MAC OSX 10.2
Thanks to Davis Thomsen for the initial document after we attacked his new
laptop.
Installing Inventor on a Ti G4 Powerbook with OSX
1) Install the developer tools that come with OSX. This can be done a
couple of ways:
a) Install from the disk that comes with OSX when bought at the store
b) Install from the packages already on the computer. This way
involves installing these packages in a specific order. In
/Applications/Installers/Developer Tools/Packages
i. install BSDSDK.pkg
ii. install DevTools.pkg
iii. finally install DevSDK.pkg
c) It is possible to download these from the apple website.
* http://connect.apple.com (can get a free login)
* Click Download Software
* Click "Developer Tools"
* December 2002 Mac OS X Developer Tools
Click "Download" for Dec 2002 Mac OS X Developer Tools
WARNING: This is over 300 MB!
2) Install X11. This can be downloaded from the apple website:
http://www.apple.com/macosx/x11/
Click Download X11 Public Beta 3
3) Install Fink and Fink Commander
Fink: http://fink.sourceforge.net/download/index.php
Fink Commander: http://finkcommander.sourceforge.net/
4) Install the X11 Developer Toolkit
At the X11 download page, http://www.apple.com/macosx/x11/download/
Near the bottom right: X11 for Mac OSX Public Beta SDK
Download SDK (4.0 Mb)
5) Use Fink Commander to download and install "system-Xfree86"
6) Use Fink Commander to download "motif" (when you type this into the
name search field, there will be three results. Install all three.
Add the library from the viz center homepage to the correct
location.
7) Use Fink Commander to download and install GMT (Generic Mapping Tool)
8) Get the Open Inventor from
http://schwehr.org/software/Darwin/OpenInventor-Works/
Untar this file and put in your HOME directory (i.e. /users/you)
Also, get the motif library on this web page and put it in
/sw/lib. This is because open inventor was build when version
2 of OpenMotif was standard in fink.
9) Add the following lines to your .cshrc:
setenv DYLD_LIBRARY_PATH ${HOME}/inventor/lib
setenv PATH ${PATH}:${HOME}/inventor/apps/tools/ivview
10) Move the "inventor" folder to ~/
11) Hopefully the software will work.
If you want to install mb_system, here are the Makefile settings that
might work for you (provided by Jasper Konter). OS X support for
MB-System should be increasing now that Dave Caress has a Titanium G4
laptop as his primary development platform.
ftp://ftp.ldeo.columbia.edu/pub/mbsystem
# **** EDIT THE PARAMETERS HERE ****
# Set the environment parameters:
$MBSYSTEM_HOME = "/sw/share/mbsystem";
$OS = "DARWIN";
$BYTESWAPPED = "NO";
$CC = "gcc";
$GRAPHICAL = "YES";
$MOTIFINCDIR = "/sw/include/Xm";
$MOTIFLIBS = "-L/sw/lib -L/usr/X11R6/lib -L/usr/X11R6/include -lXm -lXt -lX11";
$GMTVERSION = "3.4";
$GMTLIBDIR = "/sw/lib";
$GMTINCDIR = "/sw/include";
$NETCDFLIBDIR = "/sw/lib";
$NETCDFINCDIR = "/sw/include";
$LEVITUS = "/sw/share/levitus/annual";
$PROJECTIONS = "$MBSYSTEM_HOME/share/Projections.dat";
$CFLAGS = "-g -I/usr/include/X11R6 -I/usr/X11R6/include -I/sw/include -no-cpp-precomp";
$LFLAGS = "-Wl -force_flat_namespace -lm";
#
# **** LEAVE EVERYTHING ELSE BELOW ALONE ****
----------------------------------------------------------------------
* EMACS QUICK STARTUP
Emacs is a very powerful editor and I highly recommend it for all
sorts of editing and programming tasks. However, it is pretty hard to
use till you get the hang of it. Here is a quick tutorial on how to
use emacs.
The first thing to do is to setup emacs with a special configuration
file if you do not already have one. I have a sample setup that you
can use:
cd
cp ~schwehr/.emacs .
cp -r ~schwehr/.emacs-lisp .
Or here is where you can get the OpenInventor mode files yourself:
cd
mkdir .emacs-lisp
cd .emacs-lisp
wget ftp://ftp.sgi.com/sgi/inventor/emacs/iv.el
wget ftp://ftp.sgi.com/sgi/inventor/emacs/iv-mode.el
Here is the alias that I use to startup emacs in bash:
alias e='emacs -fn 6x13 -bg black -fg white'
Okay, now we are ready to see a list of commands. The first thing is
a quick note on the "meta" and "control" keys. The meta key is
written as "M-" and can be done by either pressing the Esc key once
followed by the key or by holding down "Alt" or the diamond key on
suns and pressing the key at the same time. Experiment to find out
what works on your computer. The control key is written as C- and
works just as you would expect. You can see a lot of the key
combinations if the menus are working by just by clicking on the
menu. The key equivalent is listed.
C-x C-c Exit emacs
C-x C-s Save
C-x C-w Save As a different name
C-x C-f Open a file ( hint: open "." for a directory)
C-x k Close a buffer (a.k.a. file)
C-x i Insert a file
C-g QUIT!!! Use this if stuck.
C-x 2 Split screen editing: top-bottom
C-x 3 Split screen editing: left-right
C-x 1 Single file editing
C-x b Jump to a different open file/buffer
C-s Search for string (type C-s again to find next)
C-r Search backwards
C-k Cut (Kut) to the end of the line
C-y Yank the previous cut. (M-y right after to go
back to previous cuts)
C-_ UNDO! (C-z again to keep undoing)
C-space Mark the beginning of a region
C-w Cut the region
M-w Copy the region
M-x query-replace Go through and replace one string with
another, but ask first on each one.
M-x replace-string Replace all occurrences of a string
M-x goto-line Jump to a line number
C-x C-g Jump to a line if you use my .emacs
M-x compile Compile or run a program