Fedora Project

Why you can't trust numeric computation? (Python / Julia / Octave)

There are some matrices which are just not meant for computation. They are known as ill conditioned matrices. "Condition number" is something that tells how "bad" . Luckily, this function is already present in numpy (numpy.linalg.cond). Bigger condition number means ill conditioned matrix.

Find out at-least 5 such matrices and show how they are giving garbage values. (You can, for example, take a matrix A, calculate it's inverse (numpy.linalg.inv), and multiply the inverse again with A, and check if the matrix is an identity matrix)

Task tags

  • python
  • numpy
  • matrix

Students who completed this task

paraxor, ujik, dhrug, Abee, HeyPetah, Ayush19, Santhu

Task type

  • code Code
  • assessment Outreach / Research
close

2019