Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 6 additions & 12 deletions docs/testing.md
Original file line number Diff line number Diff line change
Expand Up @@ -20,34 +20,28 @@ What you are testing: does your existing NumPy code run with numpy2x, and does i

- **Python ≥ 3.12, 64-bit.** Check with `python --version`.
- **NumPy ≥ 2.5.** It is upgraded automatically during installation if needed (see the note below).
- **A C++20 compiler.** There are no prebuilt wheels yet, so numpy2x is compiled during installation:
- Prebuilt wheels exist for Python 3.12 and 3.13 on Windows x64 and Linux x86_64. Elsewhere pip
compiles numpy2x during installation, which needs a C++20 compiler:
- **Windows:** Visual Studio 2022 or the *Build Tools for Visual Studio* with the
*Desktop development with C++* workload
- **Linux:** a recent GCC or Clang
- **git**, only for the first install command below. pip fetches CMake and Ninja itself.

pip fetches CMake and Ninja itself.

## 2. Install into your existing Python

Use the same Python installation that you (or your IDE) normally work with:

```bash
python -m pip install "numpy2x @ git+https://github.com/ILNumerics/numpy2x"
```

Without git:

```bash
python -m pip install "numpy2x @ https://github.com/ILNumerics/numpy2x/archive/refs/heads/main.zip"
python -m pip install numpy2x
```

The build takes about a minute.

> [!WARNING]
> pip upgrades NumPy in this environment to ≥ 2.5. Packages that require an older NumPy may stop working
> (for example, Numba reports *"Numba needs NumPy 2.1 or less"*). Check beforehand with:
>
> ```bash
> python -m pip install --dry-run "numpy2x @ git+https://github.com/ILNumerics/numpy2x"
> python -m pip install --dry-run numpy2x
> ```

**Safer alternative:** a separate environment that still sees all your installed packages. The new NumPy
Expand Down
Loading