How to calculate number of Trues per row in polars

In Pandas, calculating the number of Trues can easily done by the .sum() function on either a column (axis=0) or row (axis=1).

However, in polars, this only seems to work on individual columns:

Input:

s = pl.DataFrame({"a": [True, False, True], "b":[True, True, False]})
print(s)
# Number of Trues in each column (This works)
print(s.sum(axis=0))
# Number of Trues in each row (This does not work)
print(s.sum(axis=1))

Output:

shape: (3, 2)
┌───────┬───────┐
│ a     ┆ b     │
│ ---   ┆ ---   │
│ bool  ┆ bool  │
╞═══════╪═══════╡
│ true  ┆ true  │
│ false ┆ true  │
│ true  ┆ false │
└───────┴───────┘
shape: (1, 2)
┌─────┬─────┐
│ a   ┆ b   │
│ --- ┆ --- │
│ u32 ┆ u32 │
╞═════╪═════╡
│ 2   ┆ 2   │
└─────┴─────┘
---------------------------------------------------------------------------
PanicException                            Traceback (most recent call last)
Cell In[125], line 5
      2 print(s)
      3 print(s.sum(axis=0))
----> 5 s.sum(axis=1)

File c:\Users\xxxx\.venv\lib\site-packages\polars\dataframe\frame.py:7006, in DataFrame.sum(self, axis, null_strategy)
   7004     return self._from_pydf(self._df.sum())
   7005 if axis == 1:
-> 7006     return wrap_s(self._df.hsum(null_strategy))
   7007 raise ValueError("Axis should be 0 or 1.")

PanicException: `add` operation not supported for dtype `bool`

How can I achieve the calculation over the axis=1?
For non-boolean values this works, but for boolean values not.

(My polars verion is 0.16.18)

Thanks.

>Solution :

Solution: update your polars version

On polars 0.18.4 this looks correct:

In [36]: s = pl.DataFrame({"a": [True, False, True], "b":[True, True, False]})
    ...: print(s)
    ...: # Number of Trues in each column (This works)
    ...: print(s.sum(axis=0))
    ...: # Number of Trues in each row (This does not work)
    ...: print(s.sum(axis=1))
shape: (3, 2)
┌───────┬───────┐
│ a     ┆ b     │
│ ---   ┆ ---   │
│ bool  ┆ bool  │
╞═══════╪═══════╡
│ true  ┆ true  │
│ false ┆ true  │
│ true  ┆ false │
└───────┴───────┘
shape: (1, 2)
┌─────┬─────┐
│ a   ┆ b   │
│ --- ┆ --- │
│ u32 ┆ u32 │
╞═════╪═════╡
│ 2   ┆ 2   │
└─────┴─────┘
shape: (3,)
Series: 'a' [u32]
[
        2
        1
        1
]