# Isposdef and eigvals do not agree

**URL:** https://discourse.julialang.org/t/isposdef-and-eigvals-do-not-agree/118191
**Category:** Statistics
**Tags:** linearalgebra, distributions, potential-bug
**Created:** [August 14, 2024, 5:02pm UTC](https://discourse.julialang.org/t/isposdef-and-eigvals-do-not-agree/118191 "2024-08-14T17:02:19Z")
**Posts on this page:** 1
**Showing post:** 6

<div class="post-metadata">

### Author: ![fmor](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/fmor/32/211337_2.png) [@fmor](https://discourse.julialang.org/u/fmor)
#### Post date: [August 14, 2024, 6:55pm UTC](https://discourse.julialang.org/t/isposdef-and-eigvals-do-not-agree/118191/6 "2024-08-14T18:55:52Z")

</div>

First of all thank you everyone for the answers 🙂

Then in the end I also ended up to this other question (that I didn’t found when composing mine): [https://discourse.julialang.org/t/all-positive-eigen-values-but-false-in-isposdef/107393](https://discourse.julialang.org/t/all-positive-eigen-values-but-false-in-isposdef/107393) so yes the problem seems to be solvable by forcing the matrix to be Symmetric or Hermitian.

---

_[View the full topic](https://discourse.julialang.org/t/isposdef-and-eigvals-do-not-agree/118191)._