Upgrade to numpy 2.2

[claudia-dribia]

Current Behaviour

I encountered a version incompatibility issue when trying to use ydata-profiling with a project that depends on NumPy 2.2. It appears that the current version of ydata-profiling specifies a dependency constraint of numpy <2.2.

Attempting to install or use ydata-profiling with NumPy 2.2 results in dependency resolution errors.

Expected Behaviour

I would like to use ydata-profiling in conjunction with NumPy 2.2 without encountering compatibility issues.

Data Description

To reproduce:

1. Set up a Python environment with NumPy 2.2 installed
2. Attempt to install ydata-profiling
3. Observe the dependency conflict.

Code that reproduces the bug

pandas-profiling version

v4.16.1

Dependencies

numpy>=1.16.0,<2.2

OS

No response

Checklist

• There is not yet another bug report for this issue in the issue tracker
• The problem is reproducible from this bug report. This guide can help to craft a minimal bug report.
• The issue has not been resolved by the entries listed under Common Issues.

[Mevrael]

Tried to install it with uv add today.

Also got a few incompatible dependency version issues as well:

numpy<1.16.0
numpy>=2.2

seaborn<0.10.1
seaborn>=0.13

typeguard<2.13.2
typeguard>=2.14

[Val11011]

same problem here
uv add ydata_profiling
× No solution found when resolving dependencies:
╰─▶ Because only the following versions of ydata-profiling are available:
ydata-profiling==4.0.0
ydata-profiling==4.1.0
ydata-profiling==4.1.1
ydata-profiling==4.1.2
ydata-profiling==4.2.0
ydata-profiling==4.3.0
ydata-profiling==4.3.1
ydata-profiling==4.3.2
ydata-profiling==4.4.0
ydata-profiling==4.5.0
ydata-profiling==4.5.1
ydata-profiling==4.6.0
ydata-profiling==4.6.1
ydata-profiling==4.6.2
ydata-profiling==4.6.3
ydata-profiling==4.6.4
ydata-profiling==4.6.5
ydata-profiling==4.7.0
ydata-profiling==4.8.3
ydata-profiling==4.9.0
ydata-profiling==4.10.0
ydata-profiling==4.11.0
ydata-profiling==4.12.0
ydata-profiling==4.12.1
ydata-profiling==4.12.2
ydata-profiling==4.13.0
ydata-profiling==4.14.0
ydata-profiling==4.15.0
ydata-profiling==4.15.1
ydata-profiling==4.16.0
ydata-profiling==4.16.1
and ydata-profiling<=4.1.2 depends on tqdm>=4.48.2,<4.65, we can conclude that ydata-profiling<4.2.0 depends on tqdm>=4.48.2,<4.65.
And because ydata-profiling>=4.2.0,<=4.7.0 depends on seaborn>=0.10.1,<0.13, we can conclude that all of:
seaborn<0.10.1
seaborn>=0.13
, all of:
tqdm<4.48.2
tqdm>=4.65
, ydata-profiling<4.8.3 are incompatible.
And because ydata-profiling>=4.7.0,<=4.9.0 depends on numpy>=1.16.0,<2 and numpy>=1.16.0,<2.2, we can conclude that all of:
numpy<1.16.0
numpy>=2.2
, all of:
seaborn<0.10.1
seaborn>=0.13
, all of:
tqdm<4.48.2
tqdm>=4.65
, all versions of ydata-profiling are incompatible.
And because your project depends on numpy>=2.2.6 and seaborn>=0.13.2, we can conclude that your project, all of:
tqdm<4.48.2
tqdm>=4.65
, all versions of ydata-profiling are incompatible.
And because your project depends on tqdm>=4.67.1 and ydata-profiling, we can conclude that your project's requirements are unsatisfiable.
help: If you want to add the package regardless of the failed resolution, provide the --frozen flag to skip locking and syncing.

[fabclmnt]

Hi @claudia-dribia,

Thank you for opening this request.

We currently keep the NumPy version fixed to avoid incompatibilities and regressions that have occurred in the past with untested versions. While an upgrade is technically possible, we don’t have a release planned until the end of this month.

Could you please let us know if there’s a specific package you’re using that explicitly requires NumPy 2.2? That information would help us evaluate whether this is something we should prioritize before our next release.

Looking forward to your feedback!