# Velocity profile from 3d velocity space data

1. Nov 19, 2007

### jimbo_durham

Hi. I have a load of data objcts, each with velocity componants in the cartesian x,y,z directions. I would like to find a measure of the distribution of velocity (assuming it is averaged in all directions - a good assumption). Ie i would like to end up with some 2d plot with velocity along the 'x' axis, and number density, propability amplitude etc up the 'y' axis.

I have plotted the velocity data in 3d to show the distribution in x,y,z. it is attached. It appears that the velocity falls off in probability as 1/r^2 or similar. ie there are lots of objects with velocitys in the low 10s of kms-1, and fewer objects in the high 100s kms-1. I have the raw data and need find the form of the function.

Basically, from the plot in the figure, i need the density of dots(representing velocity componants) as a function of radius. How can i do this from the raw data? i think this might be a binning problem, where the bins are shells with equal volume in velocity space, and the number of 'dots' which fall into the shell gives the form of the function i am after.

any better ideas?

jimbo

#### Attached Files:

• ###### velocity_profile_3d_dist.jpg
File size:
36.5 KB
Views:
69
2. Nov 20, 2007

### jimbo_durham

solved. shell binning worked fine. alhough the volume of each shell was allowed to change, this was accounted for and the number density found as a function of radial distance from center, equating to a number density profile against actual velocity.

thanks to anyone who read this and was thinking of a solution for me

3. Nov 20, 2007

### jimbo_durham

ps if anyone wants to see the results/method if they are doing anything similar, i can post them

4. Dec 23, 2009

### AzeemA

Dear Jimbo,
may be i have more complicated problem, but i believe your example would help me much, please send me the method and results to [email address deleted]

Best Regards,
Mohamed

Last edited by a moderator: Jan 31, 2010