RuntimeWarning: All-NaN slice encountered

RuntimeWarning: All-NaN slice encountered

$ python - <<'PY'
import numpy as np
arr = np.array([np.nan, np.nan])
np.nanmin(arr)
PY
# RuntimeWarning: All-NaN slice encountered

Why this happens

This warning occurs when a reduction (e.g., np.nanmin, np.nanmax, np.nanmean) operates over an array or slice that contains only NaN values. NumPy can’t compute a meaningful numeric result for such operations.

Fix

Handle NaNs before calling reductions. You can filter NaNs or provide a fallback (e.g., np.nan_to_num or conditionals that return a default). Alternatively, use masked arrays which keep NaN masked values separate from computations.

Wrong code

import numpy as np
arr = np.array([np.nan, np.nan])
print(np.nanmin(arr))

Fixed code

import numpy as np
arr = np.array([np.nan, np.nan])
# Provide fallback
if np.all(np.isnan(arr)):
    print("no valid numbers")
else:
    print(np.nanmin(arr))

# or use masked arrays
masked = np.ma.masked_invalid(arr)
print(masked.min())