• [ ] Closes #xxxx
• [X] Tests added

When running the following example the python interpreter exists abnormally (exit code -1073741819)

The code that causes the exception is the maxnan() function in bottleneck.

This error only occurs when:

1. The data contains a dimension with length 1
2. The data is transposed
3. bottleneck is used

So I suspect that the transpose function changes the data somehow such that bottleneck can not handle it anymore.

• Calling bn.maxnan on a normal ndarray works fine
• Calling data.max() before transposing works fine
• Running the example without using bottleneck (either by not installing it or by disabling it in OPTIONS) works fine
• Running the example with len(time) > 1 works fine

What happened:

Process finished with exit code -1073741819 (0xC0000005)

What you expected to happen:

Script should run without crashing

Minimal Complete Verifiable Example:

```python from collections import OrderedDict import numpy as np import xarray as xr

xr.show_versions()

n_time = 1 # 1 : Fails, 2 : everything is fine

from xarray.core.options import OPTIONS OPTIONS["use_bottleneck"] = True # Set to False for work-around

Build some dataset

dirs = np.linspace(0,360, num=121) freqs = np.linspace(0,4,num=192) spec_data = np.random.random(size=(n_time,192,121))

dims = ('time', 'freq', 'dir') coords = OrderedDict() coords['time'] = range(n_time) coords['freq'] = freqs coords['dir'] = dirs

xdata = xr.DataArray( data=spec_data, coords=coords, dims=dims, name='Spec name', ).to_dataset()

xdata = xdata.transpose(..., "freq") # remove this line and the script will run

tm = xdata.max() print('Done') ```

Anything else we need to know?:

Uhm... it was really hard to dig this deep? :-)

Environment:

Tested with python 3.8 and 3.9 on win64. Only required packages are xarray and bottleneck. For example:

yml name: ws dependencies: - python=3.9 - xarray channels: - defaults - conda-forge

note: xarray requires pandas which installs bottleneck. To install without bottleneck pin pandas to 1.2.4

I have a piece of code runs fine on an conda environment with just xarray. But when I add numba then the code crashes

What happened:

Abnormal process termination (eg Process finished with exit code -1073741819 (0xC0000005))

What you expected to happen:

should calculate the max of the data (same as before transposing)

Minimal Complete Verifiable Example:

```python from collections import OrderedDict

import numpy as np import xarray as xr

Build some dataset

dirs = np.linspace(0,360, num=121) freqs = np.linspace(0,4,num=192) spec_data = np.random.random(size=(192,121)) data = [spec_data]

dims = ('time', 'freq', 'dir') coords = OrderedDict() coords['time'] = [0,] coords['freq'] = freqs coords['dir'] = dirs

print('constructing data-array')

xdata = xr.DataArray( data=data, coords=coords, dims=dims, name='Spec name', ).to_dataset()

print('transposing data-array')

print('getting max')

print(xdata.max()) # works fine

tdata = xdata.transpose(..., "freq")

print('getting max')

print(tdata.max()) # <==== Process finished with exit code -1073741819 (0xC0000005) print('done!') ```

Anything else we need to know?:

Running on windows 10 x64 Python 3.10 but also occurs with 3.8 and 3.9 (did not test any others)

Environment:

When I create an environment with ONLY xarray then everything works as expected:

This environment WORKS

But when I add numba then it fails:

I need to interpolate (wave) forces on a ship between headings and/or frequencies.

heading and frequency are both dimensions. force is a phase + amplitude pair.

What I would normally do is linear interpolation of the force amplitude and linear interpolation of the unwrapped phase. Interpolation of the amplitude works fine, but interpolation of the phase is troublesome because I can not unwrap the phases in two dimensions (heading and frequency) at the same time.

A solution that I can think of is storing the phase as a complex number, interpolate that, and then get the phase of the interpolated value. But calculating the angle (phase) from the interpolated values would be slow and it feels like a workaround rather than a good solution.

It would be great if I could pass a custom interpolation function to the interpolate method to use instead of scipy.interpolate.interp1d. But as far as I can see this is not (yet) an option.