# Numpy issues 2x - caught between a rock and a hard place?

**URL:** <https://forum.biobakery.org/t/numpy-issues-2x-caught-between-a-rock-and-a-hard-place/6274>\
**Category:** bioBakery workflows\
**Created:** [December 1, 2023, 9:14am UTC](https://forum.biobakery.org/t/numpy-issues-2x-caught-between-a-rock-and-a-hard-place/6274 "2023-12-01T09:14:15Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![aggregatibacter](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.biobakery.org/aggregatibacter/32/2567_2.png) [@aggregatibacter](https://forum.biobakery.org/u/aggregatibacter)\
**Post date:** [December 1, 2023, 9:14am UTC](https://forum.biobakery.org/t/numpy-issues-2x-caught-between-a-rock-and-a-hard-place/6274/1 "2023-12-01T09:14:15Z")

</div>

Hi,

I am on Ubuntu 20.04 and Python 3.8.10 and I ran into multiple issues with numpy for the wmgx workflow.

Initially, with numpy 1.24.4, I got an “AttributeError: module ‘numpy’ has no attribute ‘int’.” - this went away by downgrading to 1.20.3 and allowed me to run kneaddata without issues.

Now, however, a new error appeared in metaphlan “ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject” - it was suggested, that upgrading numpy will sort this, but this will again break the aforementioned issue.

Any ideas please? Many thanks!!

---

<div class="post-metadata">

**Author:** ![lauren.j.mciver](https://avatars.discourse-cdn.com/v4/letter/l/f05b48/32.png) [@lauren.j.mciver](https://forum.biobakery.org/u/lauren.j.mciver)\
**Post date:** [January 19, 2024, 4:34pm UTC](https://forum.biobakery.org/t/numpy-issues-2x-caught-between-a-rock-and-a-hard-place/6274/2 "2024-01-19T16:34:27Z")

</div>

Hi, Thank you for including so many details in your post! We run Python 3.10.9 with numpy 1.25.1 and this works for the latest version of Kneaddata and MetaPhlAn.

Thanks!  
Lauren