Regression Plane / Plane Fitting

markraz

markraz

Full Member
Joined
Feb 19, 2014
Messages
338
Hi, I have several hundred points in a matrix that are almost planar. Is it true that I can find the best fit plane using the following ??

x = ((A^t * A) ^-1) * A^t * b

where A is my matrix in form of:
[ [x y 1.0]
[x y 1.0]
[x y 1.0]
[x y 1.0]
[x y 1.0]
[more xy values.......]]

b are my z values

b =
[ [z]
[z]
[z]
[z]
[z]
[ more z values]]


^t = is matrix transform
^-1 = is matrix inverse
* = is multiplication


Thanks
 
I gave you a reference. I haven't looked at that stuff in ages so that's all the help I can give you. You should be able to check for yourself whether your method is the same.
 
LCKurtz said:
I gave you a reference. I haven't looked at that stuff in ages so that's all the help I can give you. You should be able to check for yourself whether your method is the same.
Click to expand...
^^^condescending ^^ watch your tone or I will report you for abuse. If I knew how to solve this problem would I be asking for help??? seriously??
I know how to use google though
 
Last edited:
markraz said:
^^^condescending ^^ watch your tone or I will report you for abuse. If I knew how to solve this problem would I be asking for help??? seriously??
I know how to use google though
Click to expand...
I certainly don't think so. In response #4, the author simply states that s/he does not have neither the time nor the "current knowledge" and/or interest to follow through and analyze different methods.
 
Subhotosh Khan said:
I certainly don't think so. In response #4, the author simply states that s/he does not have neither the time nor the "current knowledge" and/or interest to follow through and analyze different methods.
Click to expand...
Thanks for that Subhotosh. I was about to reply in a way that might have actually deserved a reprimand before I saw your post...
 
You must log in or register to reply here.
Top