e10v commented on issue #14882:
URL: https://github.com/apache/arrow/issues/14882#issuecomment-2555518906
Here is the alternative proposed by ChatGPT :)
```python
import pyarrow as pa
import pyarrow.compute as pc
# Example data
table = pa.Table.from_pydict({
"column_a": pa.array([1, 2, 1, 3, 2]),
"column_b": pa.array(["x", "y", "z", "x", "y"])
})
group_col = "column_a"
# 1. Sort indices by the grouping column
sort_indices = pc.sort_indices(table, sort_keys=[(group_col, "ascending")])
sorted_table = table.take(sort_indices)
# 2. Identify boundaries where column_a changes
col = sorted_table[group_col]
# Compare adjacent values to find boundaries:
equal_to_previous = pc.equal(col[1:], col[:-1])
changes = pc.invert(equal_to_previous) # True where value changes
# Convert boolean array to offsets
# Start at 0
offsets = [0]
# Add 1 to shift indices so they match starts of groups
for i, changed in enumerate(changes.to_pylist(), start=1):
if changed:
offsets.append(i)
# Final boundary is the length of the table
offsets.append(len(col))
# 3. Slice groups
for start, end in zip(offsets, offsets[1:]):
subtable = sorted_table.slice(start, end - start)
key = subtable[group_col][0].as_py()
...
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
