In numeric applications, all Linear Algebra concepts are conflated to a single object: The numpy array. In this library we provide a collection of sub-classes of np.array that specialize the interpretation of the array as linear map (Matrix), sub-vector space (Vect), etc. and provide a collection of common high-level operations for the concept at hand.
Disclaimer This library is in it's infancy. Expect things to change without warning.
> X = Matrix.rand_rk(10, 5, 3) # random 10x5 matrix of rank 3
> X.rank()
3
> X # pretty printing by default
Matrix[10,5]{
[ 1 -0.62 -0.16 -1.1 -2.6]
[ 0.58 0.019 1.6 1.8 -1]
[ -0.36 -0.54 -0.35 -2 0.018]
[ -0.17 -0.078 -1.5 -1.5 0.21]
[ -0.68 -0.42 1.4 -0.4 0.75]
[-0.054 -0.019 0.76 0.54 0.08]
[ 0.44 -0.5 0.032 -1 -1.4]
[ -0.36 -0.65 1 -1.2 -0.11]
[ 0.2 -0.01 0.051 0.15 -0.37]
[ 0.74 0.21 -1.6 -0.22 -1.1]
}
> X[1:3] # numpy operations keep working and yield Matrix objects
Matrix[2,5]{
[ 0.58 0.019 1.6 1.8 -1]
[ -0.36 -0.54 -0.35 -2 0.018]
}
> V = Vect.Im(X) # Image vector space
> V.dim()
3
> V.contains(X.get_col(3))
True
> V.project([1,2,3,4,5,6,7,8,9,10]).to_row()
Matrix[1,10]{
[ 6.9 0.25 3.6 0.94 1.2 -0.32 4.7 3.8 0.45 0.62]
}See ./examples for more usage examples.
pip install libla
- 2020-03-07 v0.2 Revised project structure