Sparse direct solvers (spsolve) in SciPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the purpose of the spsolve function in SciPy?

spsolve solves linear systems where the matrix is sparse. It finds the solution x for Ax = b efficiently by using direct methods designed for sparse matrices.

beginner

What type of matrix does spsolve expect as input?

spsolve expects a sparse matrix, usually in CSR (Compressed Sparse Row) or CSC (Compressed Sparse Column) format. These formats store only non-zero elements to save memory and speed up calculations.

beginner

Why use sparse direct solvers instead of dense solvers?

Sparse direct solvers like spsolve use less memory and run faster for large sparse systems because they skip zero elements. Dense solvers waste time and memory handling zeros.

intermediate

What is the difference between direct and iterative solvers?

Direct solvers find the exact solution in a finite number of steps (like spsolve). Iterative solvers start with a guess and improve it step-by-step, useful for very large or complex systems.

beginner

How do you prepare the right-hand side vector b for spsolve?

The vector b should be a 1D NumPy array or a compatible vector representing the constants in the equation Ax = b. It must match the size of the matrix A.

Which matrix format is best suited for spsolve?

A Diagonal matrix only

B Dense NumPy array

C List of lists

D CSR or CSC sparse matrix

spsolve works best with sparse matrices in CSR or CSC format to efficiently handle non-zero elements.

What does spsolve(A, b) compute?

A Inverse of matrix <code>A</code>

B Solution vector <code>x</code> for <code>Ax = b</code>

C Determinant of matrix <code>A</code>

D Eigenvalues of matrix <code>A</code>

spsolve solves the linear system Ax = b and returns x.

Why is spsolve preferred over dense solvers for large sparse systems?

A It converts sparse to dense internally

B It always gives approximate solutions

C It uses less memory and runs faster

D It only works for small matrices

spsolve is optimized for sparse matrices, saving memory and time by ignoring zeros.

What happens if you pass a dense matrix to spsolve?

A It will convert it to sparse internally but may be inefficient

B It will raise an error immediately

C It will solve faster than sparse matrices

D It will ignore the matrix and return zeros

spsolve can accept dense matrices but is designed for sparse ones; performance may suffer.

Which of these is NOT a direct solver?

A Conjugate Gradient method

B LU decomposition

C spsolve

D Cholesky decomposition

Conjugate Gradient is an iterative solver, unlike spsolve and LU or Cholesky decompositions which are direct methods.

Explain how spsolve works and why it is useful for sparse matrices.

Think about how sparse matrices have many zeros and how <code>spsolve</code> skips them.

Describe the difference between direct and iterative solvers in the context of sparse linear systems.

Consider accuracy and speed for different problem sizes.