3D plot of scattered data

Share icon

Thu, 07/31/2014 - 05:52 am

I have a matrix created from a Raman spatial mapping which I would like to plot in 3D. This matrix is created by the software I use to acquire data.
The table is pretty straightforward, the first two columns are the X, Y positions where every spectrum was collected and for every row I have the intensity at different wavenumbers.

It is almost like a scattered data, but instead of having one Z column for each XY position, I have 2654 columns related to each position. Is it possible to plot it?

Cheers
Do you an idea of what you are aiming for here? Your question is a bit vague.
In the distant past I have done things along the lines of pick particular features in the Raman spectra (say a peak area), and assign several (typically three) of these to a different colour (red, green and blue) and plot these as map. I have also recorded reference Raman spectra from pure materials and fitted each recorded 'map' spectrum (from a pharmaceutical tablet, for example) to a linear combination of these reference spectra, and then created colour maps again as described above.
Needless to say I used Igor for all this analysis, but it was some 10 to 15 years ago, or more.
Hope this gives some ideas for you to think about.
Regards,
Kurt

July 31, 2014 at 06:49 am - Permalink

Thanks Kurt!

That is precisely what I want to do with the data I have. I need to assign colours (three) to peaks and plot them as map. I also would like to plot the intensity ratio of these peaks. I've only been using Igor since last year, so I didn't know I could do that. Thanks for your help because I don't know where to start with!

Best regards,
Raquel

August 1, 2014 at 03:30 am - Permalink

raponzoni wrote:
I need to assign colours (three) to peaks and plot them as map. I also would like to plot the intensity ratio of these peaks.


Okay, there are several ways of doing this. One is do the following:
1. Make a 'results wave' that is 2654 rows by 3 (or more) columns.
2. Write a function to loop round each spectrum (wave) in turn.
2a. Fit the current spectrum (wave) with appropriate parameters to get your peak area of interest and assign this to value to column 0 of the results wave
2b. Repeat for other features in your spectrum, assigning to columns 1, 2, 3, etc.
3. Make a 3D wave ('myImageWaveName') which is rows x columns x 3, where rows x columns = 2654 (this assumes your data is recorded from a rectangular array).
4. Assign data from your results wave to this 'myImageWaveName' wave, where column 0 corresponds to red, 1 to green and 2 to blue.
5. Make an image using "NewImage myImageWaveName".
You can always reassign other columns from your results wave to the layers of the image wave to show the comparisons of interest.

Good luck!
Kurt

August 1, 2014 at 04:04 am - Permalink