polars_core/frame/row/
transpose.rs

1use std::borrow::Cow;
2
3use either::Either;
4
5use super::*;
6
7impl DataFrame {
8    pub(crate) fn transpose_from_dtype(
9        &self,
10        dtype: &DataType,
11        keep_names_as: Option<PlSmallStr>,
12        names_out: &[PlSmallStr],
13    ) -> PolarsResult<DataFrame> {
14        let new_width = self.height();
15        let new_height = self.width();
16        // Allocate space for the transposed columns, putting the "row names" first if needed
17        let mut cols_t = match keep_names_as {
18            None => Vec::<Column>::with_capacity(new_width),
19            Some(name) => {
20                let mut tmp = Vec::<Column>::with_capacity(new_width + 1);
21                tmp.push(
22                    StringChunked::from_iter_values(
23                        name,
24                        self.get_column_names_owned().into_iter(),
25                    )
26                    .into_column(),
27                );
28                tmp
29            },
30        };
31
32        let cols = self.columns();
33        match dtype {
34            #[cfg(feature = "dtype-i8")]
35            DataType::Int8 => numeric_transpose::<Int8Type>(cols, names_out, &mut cols_t),
36            #[cfg(feature = "dtype-i16")]
37            DataType::Int16 => numeric_transpose::<Int16Type>(cols, names_out, &mut cols_t),
38            DataType::Int32 => numeric_transpose::<Int32Type>(cols, names_out, &mut cols_t),
39            DataType::Int64 => numeric_transpose::<Int64Type>(cols, names_out, &mut cols_t),
40            #[cfg(feature = "dtype-u8")]
41            DataType::UInt8 => numeric_transpose::<UInt8Type>(cols, names_out, &mut cols_t),
42            #[cfg(feature = "dtype-u16")]
43            DataType::UInt16 => numeric_transpose::<UInt16Type>(cols, names_out, &mut cols_t),
44            DataType::UInt32 => numeric_transpose::<UInt32Type>(cols, names_out, &mut cols_t),
45            DataType::UInt64 => numeric_transpose::<UInt64Type>(cols, names_out, &mut cols_t),
46            DataType::Float32 => numeric_transpose::<Float32Type>(cols, names_out, &mut cols_t),
47            DataType::Float64 => numeric_transpose::<Float64Type>(cols, names_out, &mut cols_t),
48            #[cfg(feature = "object")]
49            DataType::Object(_) => {
50                // this requires to support `Object` in Series::iter which we don't yet
51                polars_bail!(InvalidOperation: "Object dtype not supported in 'transpose'")
52            },
53            _ => {
54                let phys_dtype = dtype.to_physical();
55                let mut buffers = (0..new_width)
56                    .map(|_| {
57                        let buf: AnyValueBufferTrusted = (&phys_dtype, new_height).into();
58                        buf
59                    })
60                    .collect::<Vec<_>>();
61
62                let columns = self
63                    .materialized_column_iter()
64                    // first cast to supertype before casting to physical to ensure units are correct
65                    .map(|s| s.cast(dtype)?.cast(&phys_dtype))
66                    .collect::<PolarsResult<Vec<_>>>()?;
67
68                // this is very expensive. A lot of cache misses here.
69                // This is the part that is performance critical.
70                for series in &columns {
71                    polars_ensure!(
72                        series.dtype() == &phys_dtype,
73                        ComputeError: "cannot transpose with supertype: {}", dtype
74                    );
75                    for (av, buf) in series.iter().zip(buffers.iter_mut()) {
76                        // SAFETY: we checked the type and we borrow
77                        unsafe {
78                            buf.add_unchecked_borrowed_physical(&av);
79                        }
80                    }
81                }
82                cols_t.extend(buffers.into_iter().zip(names_out).map(|(buf, name)| {
83                    // SAFETY: we are casting back to the supertype
84                    let mut s = unsafe { buf.into_series().cast_unchecked(dtype).unwrap() };
85                    s.rename(name.clone());
86                    s.into()
87                }));
88            },
89        };
90
91        DataFrame::new(new_height, cols_t)
92    }
93
94    pub fn transpose(
95        &mut self,
96        keep_names_as: Option<&str>,
97        new_col_names: Option<Either<String, Vec<String>>>,
98    ) -> PolarsResult<DataFrame> {
99        let new_col_names = match new_col_names {
100            None => None,
101            Some(Either::Left(v)) => Some(Either::Left(v.into())),
102            Some(Either::Right(v)) => Some(Either::Right(
103                v.into_iter().map(Into::into).collect::<Vec<_>>(),
104            )),
105        };
106
107        self.transpose_impl(keep_names_as, new_col_names)
108    }
109    /// Transpose a DataFrame. This is a very expensive operation.
110    pub fn transpose_impl(
111        &mut self,
112        keep_names_as: Option<&str>,
113        new_col_names: Option<Either<PlSmallStr, Vec<PlSmallStr>>>,
114    ) -> PolarsResult<DataFrame> {
115        // We must iterate columns as [`AnyValue`], so we must be contiguous.
116        self.rechunk_mut_par();
117
118        let mut df = Cow::Borrowed(self); // Can't use self because we might drop a name column
119        let names_out = match new_col_names {
120            None => (0..self.height())
121                .map(|i| format_pl_smallstr!("column_{i}"))
122                .collect(),
123            Some(cn) => match cn {
124                Either::Left(name) => {
125                    let new_names = self.column(name.as_str()).and_then(|x| x.str())?;
126                    polars_ensure!(new_names.null_count() == 0, ComputeError: "Column with new names can't have null values");
127                    df = Cow::Owned(self.drop(name.as_str())?);
128                    new_names.no_null_iter().map(PlSmallStr::from_str).collect()
129                },
130                Either::Right(names) => {
131                    polars_ensure!(names.len() == self.height(), ShapeMismatch: "Length of new column names must be the same as the row count");
132                    names
133                },
134            },
135        };
136        if let Some(cn) = keep_names_as {
137            // Check that the column name we're using for the original column names is unique before
138            // wasting time transposing
139            polars_ensure!(names_out.iter().all(|a| a.as_str() != cn), Duplicate: "{} is already in output column names", cn)
140        }
141        polars_ensure!(
142            df.height() != 0 && df.width() != 0,
143            NoData: "unable to transpose an empty DataFrame"
144        );
145        let dtype = df.get_supertype().unwrap()?;
146        df.transpose_from_dtype(&dtype, keep_names_as.map(PlSmallStr::from_str), &names_out)
147    }
148}
149
150#[inline]
151unsafe fn add_value<T: NumericNative>(
152    values_buf_ptr: usize,
153    col_idx: usize,
154    row_idx: usize,
155    value: T,
156) {
157    let vec_ref: &mut Vec<Vec<T>> = &mut *(values_buf_ptr as *mut Vec<Vec<T>>);
158    let column = vec_ref.get_unchecked_mut(col_idx);
159    let el_ptr = column.as_mut_ptr();
160    *el_ptr.add(row_idx) = value;
161}
162
163// This just fills a pre-allocated mutable series vector, which may have a name column.
164// Nothing is returned and the actual DataFrame is constructed above.
165pub(super) fn numeric_transpose<T: PolarsNumericType>(
166    cols: &[Column],
167    names_out: &[PlSmallStr],
168    cols_t: &mut Vec<Column>,
169) {
170    let new_width = cols[0].len();
171    let new_height = cols.len();
172
173    let has_nulls = cols.iter().any(|s| s.null_count() > 0);
174
175    let mut values_buf: Vec<Vec<T::Native>> = (0..new_width)
176        .map(|_| Vec::with_capacity(new_height))
177        .collect();
178    let mut validity_buf: Vec<_> = if has_nulls {
179        // we first use bools instead of bits, because we can access these in parallel without aliasing
180        (0..new_width).map(|_| vec![true; new_height]).collect()
181    } else {
182        (0..new_width).map(|_| vec![]).collect()
183    };
184
185    // work with *mut pointers because we it is UB write to &refs.
186    let values_buf_ptr = &mut values_buf as *mut Vec<Vec<T::Native>> as usize;
187    let validity_buf_ptr = &mut validity_buf as *mut Vec<Vec<bool>> as usize;
188
189    RAYON.install(|| {
190        cols.iter()
191            .map(Column::as_materialized_series)
192            .enumerate()
193            .for_each(|(row_idx, s)| {
194                let s = s.cast(&T::get_static_dtype()).unwrap();
195                let ca = s.unpack::<T>().unwrap();
196
197                // SAFETY:
198                // we access in parallel, but every access is unique, so we don't break aliasing rules
199                // we also ensured we allocated enough memory, so we never reallocate and thus
200                // the pointers remain valid.
201                if has_nulls {
202                    for (col_idx, opt_v) in ca.iter().enumerate() {
203                        match opt_v {
204                            None => unsafe {
205                                let validity_vec: &mut Vec<Vec<bool>> =
206                                    &mut *(validity_buf_ptr as *mut Vec<Vec<bool>>);
207                                let column = validity_vec.get_unchecked_mut(col_idx);
208                                let el_ptr = column.as_mut_ptr();
209                                *el_ptr.add(row_idx) = false;
210                                // we must initialize this memory otherwise downstream code
211                                // might access uninitialized memory when the masked out values
212                                // are changed.
213                                add_value(values_buf_ptr, col_idx, row_idx, T::Native::default());
214                            },
215                            Some(v) => unsafe {
216                                add_value(values_buf_ptr, col_idx, row_idx, v);
217                            },
218                        }
219                    }
220                } else {
221                    for (col_idx, v) in ca.into_no_null_iter().enumerate() {
222                        unsafe {
223                            let column: &mut Vec<Vec<T::Native>> =
224                                &mut *(values_buf_ptr as *mut Vec<Vec<T::Native>>);
225                            let el_ptr = column.get_unchecked_mut(col_idx).as_mut_ptr();
226                            *el_ptr.add(row_idx) = v;
227                        }
228                    }
229                }
230            })
231    });
232
233    let par_iter = values_buf
234        .into_par_iter()
235        .zip(validity_buf)
236        .zip(names_out)
237        .map(|((mut values, validity), name)| {
238            // SAFETY:
239            // all values are written we can now set len
240            unsafe {
241                values.set_len(new_height);
242            }
243
244            let validity = if has_nulls {
245                let validity = Bitmap::from_trusted_len_iter(validity.iter().copied());
246                if validity.unset_bits() > 0 {
247                    Some(validity)
248                } else {
249                    None
250                }
251            } else {
252                None
253            };
254
255            let arr = PrimitiveArray::<T::Native>::new(
256                T::get_static_dtype().to_arrow(CompatLevel::newest()),
257                values.into(),
258                validity,
259            );
260            ChunkedArray::<T>::with_chunk(name.clone(), arr).into_column()
261        });
262    RAYON.install(|| cols_t.par_extend(par_iter));
263}
264
265#[cfg(test)]
266mod test {
267    use super::*;
268
269    #[test]
270    fn test_transpose() -> PolarsResult<()> {
271        let mut df = df![
272            "a" => [1, 2, 3],
273            "b" => [10, 20, 30],
274        ]?;
275
276        let out = df.transpose(None, None)?;
277        let expected = df![
278            "column_0" => [1, 10],
279            "column_1" => [2, 20],
280            "column_2" => [3, 30],
281
282        ]?;
283        assert!(out.equals_missing(&expected));
284
285        let mut df = df![
286            "a" => [Some(1), None, Some(3)],
287            "b" => [Some(10), Some(20), None],
288        ]?;
289        let out = df.transpose(None, None)?;
290        let expected = df![
291            "column_0" => [1, 10],
292            "column_1" => [None, Some(20)],
293            "column_2" => [Some(3), None],
294
295        ]?;
296        assert!(out.equals_missing(&expected));
297
298        let mut df = df![
299            "a" => ["a", "b", "c"],
300            "b" => [Some(10), Some(20), None],
301        ]?;
302        let out = df.transpose(None, None)?;
303        let expected = df![
304            "column_0" => ["a", "10"],
305            "column_1" => ["b", "20"],
306            "column_2" => [Some("c"), None],
307
308        ]?;
309        assert!(out.equals_missing(&expected));
310        Ok(())
311    }
312}
polars_core/frame/row/transpose.rs

polars_core/frame/row/
transpose.rs
