Why does numpy.linalg.solve() offer more precise matrix inversions than numpy.linalg.inv()?

by

np.linalg.solve(A, b) does not compute the inverse of A. Instead it calls one of the gesv LAPACK routines, which first factorizes A using LU decomposition, then solves for x using forward and backward substitution (see here).

np.linalg.inv uses the same method to compute the inverse of A by solving for A-1 in A·A-1 = I where I is the identity*. The factorization step is exactly the same as above, but it takes more floating point operations to solve for A-1 (an n×n matrix) than for x (an n-long vector). Additionally, if you then wanted to obtain x via the identity A-1·b = x then the extra matrix multiplication would incur yet more floating point operations, and therefore slower performance and more numerical error.

There’s no need for the intermediate step of computing A-1 – it is faster and more accurate to obtain x directly.

* The relevant bit of source for inv is here. Unfortunately it’s a bit tricky to understand since it’s templated C. The important thing to note is that an identity matrix is being passed to the LAPACK solver as parameter B.

More Related Contents:

  1. What are the differences between numpy arrays and matrices? Which one should I use?
  2. numpy matrix vector multiplication [duplicate]
  3. Numpy matrix to array
  4. how does multiplication differ for NumPy Matrix vs Array classes?
  5. Convert a 1D array to a 2D array in numpy
  6. how to perform max/mean pooling on a 2d array using numpy
  7. NumPy array/matrix of mixed types
  8. How to find linearly independent rows from a matrix
  9. Convert row vector to column vector in NumPy
  10. Numpy select rows based on condition
  11. Efficient dot products of large memory-mapped arrays
  12. How to access the ith column of a NumPy multidimensional array?
  13. Python/NumPy first occurrence of subarray
  14. Concatenating two one-dimensional NumPy arrays
  15. Replacing Numpy elements if condition is met
  16. How to make a multidimension numpy array with a varying row size?
  17. How can I add new dimensions to a Numpy array?
  18. Force NumPy ndarray to take ownership of its memory in Cython
  19. How to normalize a NumPy array to within a certain range?
  20. subsampling every nth entry in a numpy array
  21. Check if two 3D numpy arrays contain overlapping 2D arrays
  22. Deprecation status of the NumPy matrix class
  23. Numpy: Creating a complex array from 2 real ones?
  24. Iterating over Numpy matrix rows to apply a function each?
  25. Repeat each values of an array different times
  26. How to create 2d array with numpy random.choice for every rows?
  27. What is the difference between np.array() and np.asarray()?
  28. Equivalent of `polyfit` for a 2D polynomial in Python
  29. Partition array into N chunks with Numpy
  30. Numpy Vector (N,1) dimension -> (N,) dimension conversion

Leave a Comment