Find the axis (e.g., rows, columns) indices of a multi-dimensional array associated with the maximum or minimum value.
```python
import numpy as np
# generate random data
x = np.random.random((3, 3, 3, 3))
minidx = x.argmin() # get min index of flattened array
maxidx = x.argmax() # get max index of flattened array
x.reshape(-1, 1)[minidx] # get min value
x.min()
x.reshape(-1, 1)[maxidx] # get max value
x.max()
# get indices for each axis
mi = np.unravel_index(minidx, x.shape)
ma = np.unravel_index(maxidx, x.shape)
# use indices to retrieve values
x[mi[0], mi[1], mi[2], mi[3]]
x[ma[0], ma[1], ma[2], ma[3]]
```
# References
- https://stackoverflow.com/questions/11332205/find-row-or-column-containing-maximum-value-in-numpy-array