RPC Edit
Release date = 2006-05-28
perl gus_rpc_edit.pl
Editor for RPC-III road load data files
This editor works a bit differently from others which you may have used. In this it saves much time and effort. But you'll have to approach it on its own terms. I've tried to make it as intuitive as possible. So before you devote yourself to studying any of what I've written below, may I suggest you just start the editor and play around with it for a while? You will find that it has most of the expected features for RLD files in RPC-III format.
You may read in MTS RPC time history files in all their variants: *.tim, *.edt, *.rsp, *.drv, *.des, *.tex, etc. You may also read in Comma Separated Value (*.csv) files provided that they adhere to the IETF standard for CSV.
Any *.csv files not complying with the IETF standard for CSV must be rectified prior to read-in. Among those not complying are CSV files from Pimento Software as output by their Roadrunner hardware. These are easily rectified, however. All you need do is pass them through the included Perl script gus_fix_pimento_csv.pl which can perform the needed adjustments.
You may write out MTS RPC time history files in binary or ASCII (TAB delimited) formats. Writing out CSV files is not supported.
You may reduce data by peak slice or by vector envelope. The best way to learn how these work is to pick any file, edit it both ways and then scrutinize the graphs and/or perform other analyses. The GUI is fairly intuitive, so best that you just try both and see.
You can batch edit any number of files all the same way. I will describe this in more detail at a later date. Find it in the pull-down menu labeled options.
Interpolates a list of arbitrary peaks and valleys. The input file format is ASCII (*.dat) with a head and body. The header contains minimal RPC header paramters (CHANNELS = 1, UNITS.CHAN_1 = N, FRAMES = 1, DELTA_T = 2.441406E-03) one per row. The body contains rows of tab delimited columns. Each row represents simultaneous TAB delimited channel points. Waveform will be monotonic of semi-arbitrary frequency so as to occupy the designated number of frames. Output is as a binary RPC Time History file. If monotonic results prove difficult to iterate, use the Pythagorized peak/slice peak slice feature to bias the time axis, horizontally streatching out and compressing peaks according to their amplitude. Refer to the example file http://starling.us/tet/gus_perl/gus_rpc_edit_pl/rpc_files/example_arb_wav.dat
Interpolates a list of peak/valley blocks with exacting control of frequency. The input file format is ASCII (*.dat) with head and body very like that described above but somewhat more complicated. Frequency control is afforded by the addition of a Hertz-Cycles channel. For instance, a block cycle *.dat file designed to output a 3-channel RPC file would itself contain four channels. This forth channel requires its own parameter (UNITS.CHAN_4 = Hz-Cycles) in the head and its own column in each row of the body. Refer to the example file http://starling.us/tet/gus_perl/gus_rpc_edit_pl/rpc_files/example_block_cycle.dat
Output as PNG files, graphs are user-configurable as to channels, X and Y scales, background color, etc. Graphs are each accompanied by a text file listing channel max/min data, etc. The Show button will send any just-created graph to the system web browser for viewing. A serious shortcoming however is that you can not yet superimpose two files, nor display them as 3D ball-of-string vector diagrams. For that refer to the section on 3rd party softwares near the end of this document.
This program is free software and a perpetual beta-release besides. That is to say, I work on improving it continually so that it is never stable. Know therfor that it carries absolutely no warranties or guarantees of any kind (expressed, implied, or even vaguely hinted at). Use at own risk and back up your data. Before playing out in the lab any files output by this editor, analyse them thoroughly to be sure you are personally satisfied with their contents. That's what I do.
Install these into Perl via ActiveState PPM, NetBSD pkgsrc or CPAN as appropriate for your OS: Text::CSV::Simple, Tk, Tk::Pane, Tk::DirTree, GD, GD::Graph and Image::Magick
Pre-configured with these defaults: MSIE when OS = Win32; Mozilla when OS = NetBSD. Other OS's not tested. Send email to make a recommendation.
Nothing here presentes itself as a difficulty. Required external dependencies are available from the expected sources: /usr/pkgsrc for NetBSD and/or CPAN for Perl. I can't say for Linux, since I have yet to try it. If any report a problem for Linux, I'd very like to help work it out and document the solution here.
No especial problems here either, except that some things are not where most folks expect to find them.
The Perl Image::Magick module is not maintained by ActiveState. Instead it is packaged together with the free external software ImageMagick downloadable from http://imagemagick.org/www/download.html?
The GD and GD::Graph modules are not maintained by ActiveState. They say it is too difficult and complicated. Other kind folks have graciously ammended this lack. Type like so into ActiveState's PPM...
ppm> help repository
ppm> repository add ``Lincoln Stein'' http://stein.cshl.org/ppm
ppm> repository add ``Randy Kobes'' http://theoryx5.uwinnipeg.ca/ppms
OpenDX is a data visualization program originally from IBM but now gone open source. Whatever it is you want to see by way of visualizing your data, however you may want to see it, OpenDX is up to the task. Its complexity is intimidating at first glance. Also it is a memory hog beyond all else that you may ever have used before. But what it can show and how it can show it is simply not to be done without. Nothing else at all compares. Soon I'll devote a separate howto for OpenDX to present in conjunction with this very RPC3 editor of mine. Here is the URL: <http://opendx.org>
A very, very distant second place indeed to OpenDX is DPlot. Actually, I was fairly happy with DPlot until I discovered OpenDX. DPlot can do some quite pretty 3D scatter plots provided that the data are none too very complicated. There is even a free viewer program. At under $40 it is well worth the cost even if it does only work on Win32. Here is the URL for DPlot <http://www.dplot.com/>
Gan Uesli Starling <gan@starling.us>
Copyright (c) 2003, 2004, 2005, 2006 Gan Uesli Starling. All rights reserved. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.