Using the "select" tool to examine collect_stars output

Richard Knowles
Apr 30, 2002

Tom Droege recently announced a program he has released to compile star listings into groups of measurements. This program takes the calibrated star files from the Michael Richmond's pipeline, and collects them into sets of measurements.

As I manually reviewed the event files and making individual plots as I went, it hit me that there "has to be a better way" to speed this process up. As a result I began working on code to accomplish this.

I wanted to be able to step forward through the file, trigger a plot with the minimum number of keystrokes and to save interesting groups of measurements to an external file. Over the past couple weeks, and with Tom's help as a guinea pig, the program has been improved to the point where it is usable (but not perfected I'm afraid).

I know a number of people are working on datasets, and so I've decided to release this code for folks to play with as the process of improvement continues.

This first version should be considered "alpha" code, and in some areas is a little flakey. Nevertheless, it will definitely speed up the process of reviewing star files, and hopefully make things less tedious. It can be found at:

http://a188-l009.rit.edu/tass/software/select/

There are two versions of this program depending upon platform differences:

		selectcy.pl		Cygwin/Wgnuplot version
		selectun.pl		LINUX version (successfully used on 
                                  Mandrake 7.2 and 8.2 systems).

As with any perl script, the programs can be invoked by:

		$ selectun.pl

The command set is pretty minimal:

		a 		Advance (forward) to a specific Star ID
		n 		Show the next set of measurements
				Show the next set of measurements
		p 		Plot the current set of measurements
		s 		Save the current set of measurements
		q 		quit
		h 		show some help

I've discovered on LINUX, after the plot has been displayed, it can be closed by typing 'q', this will return you to the command prompt for your next command. Wgnuplot however isn't as accomodating and you'll have to kill the plot window to return to the prompt.

Some further features planned are:

Save the plot to an external file with a .png extension (next item to be done)
Move backwards in the measurements file
Figure out how to deal with plotting small sets of data separated by large blank periods i.e. two nights measurements separated by a week of nothing. How to plot this such that you don't have two small blips at ether end of a blank diagram.
Clean up behavior near the end of the file and with shorter star files and extractions
Change datasets mid stream
Fix default paths to not need pre-configuration
Write up some comprehensive documentation

Comments and suggestions can be sent to me at mailto:rknowles@woh.rr.com and are welcome.

I hope this is a useful tool.

Using the "select" tool to examine collect_stars output

Richard Knowles Apr 30, 2002

Richard Knowles
Apr 30, 2002