Page 1 of 1

Statistical Analysis of comapred objects

Posted: Wed Aug 09, 2017 1:19 pm
by ziarehmann
Hello CC community,

I am a new user of CC and must say that its a very usefull tool. I have a very basic question. I want to compare two obejects in order to ensure that my design (autocad object) and 3D printed entity are similar. I am able to import both in CC and compute the distances. My question would be how statistically i can decide whether the object printed is say 98% of the actual model. After comparing both in CC, visually i can say that whether the print is good or not (by looking at the scaler field with the distances highlighted in different colours). But how to decide statitically?What are the builtin tools in CC or even externally what tools would be helpfull to statistically analyze the results of CC. Am even not able to find any examples or tutorials made on this aspect.

Hope to have some advice from the experts here!
Regards!

Re: Statistical Analysis of comapred objects

Posted: Wed Aug 09, 2017 8:22 pm
by daniel
Two simple ways:
- look at the standard deviation that is output in the Console once the distances are computed. You can apply the rule of thumb of the Normal distribution (even though your distances may not be exactly following a Normal distribution it's generally ok): 66% of the points are < 1 sigma, 95% < 2 sigmas and 99% < 3 sigmas.

- the other way is too select the cloud on which the distances where computed and start the 'Edit > SF > Histogram' tool. This will show display the distances histogram. You can click anywhere on it and you'll see a red bar with the percentage of points below