Matlab values as sage vector or matrix object ?

Cheatboy2
31 ●5

Hi all !

I tried Sage and I found that having a matlab interface is very useful, however I can't import matlab matrices as sage matrices (or numpy arrays) for further work into sage workspace.

For example :

a = matlab.linspace(1,10,20) # ok

a * 2 # ok

plot(a) # Attribute Error

vector(a) # Not Implemented Error

a.sage() # --> unable to parse output

Is there a way to avoid this problem ? Did I miss something in the documentation or does the MatlabElement type not support the same operations as the vector type does (excepted arithmetic operators) ?

Thanks !

add a comment

2 Answers

Sort by » oldest newest most voted

answered 13 years ago

Cheatboy2
31 ●5

It's ok !

It an error of my own.

Sage does provide a __matrix__ interface for MatlabElement (but not a __vector__ one so that explains my initial problem).

However it is based on string parsing so could be slow with large matrices. I gave a glance to Matlab binding package for the R programming language and it seems that the "fastest" way to communicate with matlab is to use the filesystem through mat files. Fortunately, it is easy to do in sage with scipy.sio.loadmat function and very fast.

Thanks for your answer Jason. Have a good day.