Hey Chad the call form numpy is
numpy.linalg.eigh so basically the clas linalg as a definition to calculate the eigenvectors if you pass a covariance matrix. Not sure how to pass a covariance matrix though. depends wheather you need to find the On Mar 25, 2:47 am, Chad Dombrova <[email protected]> wrote: > i understand what you're saying and i would be very interested in this as > well. what are the building blocks required? break it down for me: what's > the name of the function in numpy? what are it's inputs and outputs? > > numpy.somefunc( someval ) -> eigen_vector ?? > > once this is clear we can start figuring out how to gather your required > inputs. > > -chad > > On Mar 24, 2010, at 4:08 PM, maurizio1974 wrote: > > > No Yilmaz ,I don't want to stick a Locator on a vertex, this is what I > > am trying to do. > > What Eigen Vectors can do is to get the orientation of a volume . > > SO basically I could get the polygon orientation of the whole object > > based on the volume that will need to broken down > > to matrices of the vertex that make it to be fed to Numpy . > > Assuming that you have a cube in space and you rotate it to an > > arbitrary rotation value. > > At this point if you freeze the transformation and center the pivot > > you will have the pivot point > > oriented to the world and not the actual roattion of the cube. > > Eigen Vectors and Eigen values can give yo that rotation to be fed in > > to the transform matrix and regain that orientation based on the > > object and not the pivot point. > > This is a case but many times you can get objects form different > > places or different conversion application and need to get a decent > > pivot to manipulate the object. > > > I hope my explanation was clear enough. > > > Thank you > > > On Mar 24, 12:04 pm, Ozgur Yılmaz <[email protected]> wrote: > >> so you want to stick a locator on to an objects vertex ? > > >> E.Ozgur Yilmaz > >> Lead Technical Directorwww.ozgurfx.com > > >> On Mon, Mar 22, 2010 at 1:05 AM, maurizio1974 > >> <[email protected]>wrote: > > >>> Hi > > >>> When I was at ILM there was a cool tool that would calculate the > >>> orientation of an object based on the volume > >>> thanks to the eigenvector. > >>> There they were using some old module called Linear Algebra to get the > >>> eigenvector result from the arrays > >>> of the object, but I know that Numpy has integrated the Linear Algebra > >>> Module so I guess that is the same. > > >>> I haven't been able to find a good way to just get an object parsing > >>> on the matrices of the individual vertex and then > >>> feed a 4x4 matrix to Numpy so that would give me a matrix that applied > >>> to a Locator would result to the right orientation of the object even > >>> if the original pivot is completely in a different position like when > >>> we freeze the transform on a rotated object in space. > > >>> Anyone got a clue of a reasonably simple way to do that ? > > >>> Thank you in advance. > > >>> -- > >>>http://groups.google.com/group/python_inside_maya > > >>> To unsubscribe from this group, send email to python_inside_maya+ > >>> unsubscribegooglegroups.com or reply to this email with the words "REMOVE > >>> ME" as the subject. > > > -- > >http://groups.google.com/group/python_inside_maya > > > To unsubscribe from this group, send email to > > python_inside_maya+unsubscribegooglegroups.com or reply to this email with > > the words "REMOVE ME" as the subject. > > -- http://groups.google.com/group/python_inside_maya To unsubscribe from this group, send email to python_inside_maya+unsubscribegooglegroups.com or reply to this email with the words "REMOVE ME" as the subject.
