Skip to content

Row**s** or column**s** scaling #236

Description

@syarra

PySparse has the ability to scale multiple rows/columns via the *= operand

from pysparse.sparse import PysparseMatrix as psp

J = psp(nrow=3, ncol=3)
J[0,0] = 1
J[1,1] = 2
J[2,2] = 3

# Scale two first rows:"
J[:2,:] *= -1.0

It is not possible to do that with CySparse. The same operation fails with error message:

TypeError: unsupported operand type(s) for *=: 'cysparse.sparse.ll_mat_views.ll_mat_view_INT64_t_FLOAT64_t.LLSparseMatrixView_INT64_t_FLOAT64_t' and 'float'

Is there a way to efficiently do the same operation? (without having to call J.row_scale multiple times)

@counterclocker

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions