How to create contour plots for Ocean Optics HR2000 spectrometers with the OOILVD driver package?

[NewmanCID]

I have recently acquired an Ocean Optics HR2000 spectrometer and the OOILVD driver package.  This package is initially designed for signal acquisition in 2D (wavelength vs. intensity).  However, I would need to obtain these data as functions of time as well (e.g. a countour plot where x-axis = Time, y-axis = Wavelength, and z-axis = Intensity).  Any ideas as to how I can do this?

[Conseils]

Sounds like a job for the Intensity graph.

Check out the examples with the keyword intensity.

[kehander]

I was trying to do exactly that not so long ago.

Unfortunately, I too hit something of a dead end... An Intensity Graph's X and Y values are defined by the indicies of the Z matrix you wire to it. I never did find a way around the problem.

I was actually experimenting with a 3D surface graph at first, which I imagine you might be able to rotate to produce a similar intensity plot, but that stubbornly refused to work at the time as well.

[Conseils]

How about the contour plot vi...... ?

Search the examples with the keyword 'contour' (Labview 7.1)

"Depicts an array or real values on a contour plot, with the data arranged on a rectangular grid."

[Conseils]

 I understand now... It means that you have to scale the max / min of your data set to fit the array that will be sent into the countour plot ?
Have I understood the problem correctly?

[kehander]

I'm not at my LabView workstation at the moment, but a contour plot looks rather different from an intensity plot, doesn't it? Or is there some way to reformat the plot so it looks more like an intensity plot?

[Conseils]

You are absolutely right, in that a contour plot is lines, as found on a map.

There is a Labview package available which I looked at previously, I never went any further in progressing it at the time as I made the intensity graph do the job for me. It gave me the visualisation that I needed.

The contour graph module information is here
"Contour v2.5 produces classic level contouring in the LabVIEW environment. Diagrams are included so that Contour may be installed on any LabVIEW platform. New features in Version 2.5 include: displaying contour levels on plots, enormous performance improvement, overlays of intensity and contours, and increased data set size. Contour comes with documentation in HTML format and a full-feature example VI to get you going fast. Contour is being used in acoustic profiling, thermography, cardiology, and amplifier noise figures."
The link for the above takes you to here
http://www.kirinos.com/

Check out the picture here
The package is quoted at 95 US Dollars. If you really need it, don't have the time to do it yourself and it works, it sounds like a fair price.

You can request a copy of the algorithm here but note its in FORTRAN 4 and also for an IBM 360  that it can't be exported outside the US due to export restrictions, its 121 USD.
"The graphical presentation of experimentally or theoretically generated data sets frequently involves the construction of contour plots. A general computer algorithm has been developed for the construction of contour plots. The Contour Plot Algorithm provides for efficient and accurate contouring with a modular approach which allows flexibility in modifying the algorithm for special applications.

The algorithm accepts as input data values at a set of points irregularly distributed over a plane. The algorithm is based on an interpolation scheme in which the points in the plane are connected by straight line segments to form a set of triangles.

In general, the data is smoothed using a least-squares-error fit of the data to a bivariate polynomial. To construct the contours, interpolation along the edges of the triangles is performed, using the bivariable polynomial if data smoothing was performed. Once the contour points have been located, the contour may be drawn.

---

Contour Plot Algorithm carries the NASA case number ARC-11441. It was originally released as part of the NASA COSMIC collection."

There are probably others as well.....

[Ben]

The principles used in this example (reply #7)

http://forums.ni.com/ni/board/message?board.id=170&message.id=143663&jump=true

demonstrates how to add plots to a 3d graph.

Does that help?

Ben

[Conseils]

Thanks for sharing that with us.
I will follow the link and look at how the algorithim is implemented. It's a better form of what I was using.

[NewmanCID]

Thanks for all the insightful and helpful input!  With your advice I was able to make an intensity chart of the data (x axis = time, y axis = wavelength, and intensity in time at each wavelength in false-color).  However, because of the sheer magnitude of data (2048 pixels worth for every time increment), there was only enough memory for visualizing about 2 seconds worth of data - not very good for monitoring a separation.  Moreover, writing these data to file (10 seconds of data at ~ 10 Hz), yields about 500 pages of ASCII text (1-2 Mb).  Alas, I have decided to simultaneously pick off a handful of individual wavelengths to monitor the process instead.  This is considerably less taxing of memory and processing time and file sizes are much more manageable.  Thanks again for all of your help!