polars_core/chunked_array/ops/
row_encode.rs

1use arrow::compute::utils::combine_validities_and_many;
2use polars_row::{
3    convert_columns, RowEncodingCategoricalContext, RowEncodingContext, RowEncodingOptions,
4    RowsEncoded,
5};
6use polars_utils::itertools::Itertools;
7use rayon::prelude::*;
8
9use crate::prelude::*;
10use crate::utils::_split_offsets;
11use crate::POOL;
12
13pub fn encode_rows_vertical_par_unordered(by: &[Column]) -> PolarsResult<BinaryOffsetChunked> {
14    let n_threads = POOL.current_num_threads();
15    let len = by[0].len();
16    let splits = _split_offsets(len, n_threads);
17
18    let chunks = splits.into_par_iter().map(|(offset, len)| {
19        let sliced = by
20            .iter()
21            .map(|s| s.slice(offset as i64, len))
22            .collect::<Vec<_>>();
23        let rows = _get_rows_encoded_unordered(&sliced)?;
24        Ok(rows.into_array())
25    });
26    let chunks = POOL.install(|| chunks.collect::<PolarsResult<Vec<_>>>());
27
28    Ok(BinaryOffsetChunked::from_chunk_iter(
29        PlSmallStr::EMPTY,
30        chunks?,
31    ))
32}
33
34// Almost the same but broadcast nulls to the row-encoded array.
35pub fn encode_rows_vertical_par_unordered_broadcast_nulls(
36    by: &[Column],
37) -> PolarsResult<BinaryOffsetChunked> {
38    let n_threads = POOL.current_num_threads();
39    let len = by[0].len();
40    let splits = _split_offsets(len, n_threads);
41
42    let chunks = splits.into_par_iter().map(|(offset, len)| {
43        let sliced = by
44            .iter()
45            .map(|s| s.slice(offset as i64, len))
46            .collect::<Vec<_>>();
47        let rows = _get_rows_encoded_unordered(&sliced)?;
48
49        let validities = sliced
50            .iter()
51            .flat_map(|s| {
52                let s = s.rechunk();
53                #[allow(clippy::unnecessary_to_owned)]
54                s.as_materialized_series()
55                    .chunks()
56                    .to_vec()
57                    .into_iter()
58                    .map(|arr| arr.validity().cloned())
59            })
60            .collect::<Vec<_>>();
61
62        let validity = combine_validities_and_many(&validities);
63        Ok(rows.into_array().with_validity_typed(validity))
64    });
65    let chunks = POOL.install(|| chunks.collect::<PolarsResult<Vec<_>>>());
66
67    Ok(BinaryOffsetChunked::from_chunk_iter(
68        PlSmallStr::EMPTY,
69        chunks?,
70    ))
71}
72
73/// Get the [`RowEncodingContext`] for a certain [`DataType`].
74///
75/// This should be given the logical type in order to communicate Polars datatype information down
76/// into the row encoding / decoding.
77pub fn get_row_encoding_context(dtype: &DataType) -> Option<RowEncodingContext> {
78    match dtype {
79        DataType::Boolean
80        | DataType::UInt8
81        | DataType::UInt16
82        | DataType::UInt32
83        | DataType::UInt64
84        | DataType::Int8
85        | DataType::Int16
86        | DataType::Int32
87        | DataType::Int64
88        | DataType::Int128
89        | DataType::Float32
90        | DataType::Float64
91        | DataType::String
92        | DataType::Binary
93        | DataType::BinaryOffset
94        | DataType::Null
95        | DataType::Time
96        | DataType::Date
97        | DataType::Datetime(_, _)
98        | DataType::Duration(_) => None,
99
100        DataType::Unknown(_) => panic!("Unsupported in row encoding"),
101
102        #[cfg(feature = "object")]
103        DataType::Object(_, _) => panic!("Unsupported in row encoding"),
104
105        #[cfg(feature = "dtype-decimal")]
106        DataType::Decimal(precision, _) => {
107            Some(RowEncodingContext::Decimal(precision.unwrap_or(38)))
108        },
109
110        #[cfg(feature = "dtype-array")]
111        DataType::Array(dtype, _) => get_row_encoding_context(dtype),
112        DataType::List(dtype) => get_row_encoding_context(dtype),
113        #[cfg(feature = "dtype-categorical")]
114        DataType::Categorical(revmap, ordering) | DataType::Enum(revmap, ordering) => {
115            let revmap = revmap.as_ref().unwrap();
116
117            let (num_known_categories, lexical_sort_idxs) = match revmap.as_ref() {
118                RevMapping::Global(map, _, _) => {
119                    let num_known_categories = map.keys().max().copied().map_or(0, |m| m + 1);
120
121                    // @TODO: This should probably be cached.
122                    let lexical_sort_idxs =
123                        matches!(ordering, CategoricalOrdering::Lexical).then(|| {
124                            let read_map = crate::STRING_CACHE.read_map();
125                            let payloads = read_map.get_current_payloads();
126                            assert!(payloads.len() >= num_known_categories as usize);
127
128                            let mut idxs = (0..num_known_categories).collect::<Vec<u32>>();
129                            idxs.sort_by_key(|&k| payloads[k as usize].as_str());
130                            let mut sort_idxs = vec![0; num_known_categories as usize];
131                            for (i, idx) in idxs.into_iter().enumerate_u32() {
132                                sort_idxs[idx as usize] = i;
133                            }
134                            sort_idxs
135                        });
136
137                    (num_known_categories, lexical_sort_idxs)
138                },
139                RevMapping::Local(values, _) => {
140                    // @TODO: This should probably be cached.
141                    let lexical_sort_idxs =
142                        matches!(ordering, CategoricalOrdering::Lexical).then(|| {
143                            assert_eq!(values.null_count(), 0);
144                            let values: Vec<&str> = values.values_iter().collect();
145
146                            let mut idxs = (0..values.len() as u32).collect::<Vec<u32>>();
147                            idxs.sort_by_key(|&k| values[k as usize]);
148                            let mut sort_idxs = vec![0; values.len()];
149                            for (i, idx) in idxs.into_iter().enumerate_u32() {
150                                sort_idxs[idx as usize] = i;
151                            }
152                            sort_idxs
153                        });
154
155                    (values.len() as u32, lexical_sort_idxs)
156                },
157            };
158
159            let ctx = RowEncodingCategoricalContext {
160                num_known_categories,
161                is_enum: matches!(dtype, DataType::Enum(_, _)),
162                lexical_sort_idxs,
163            };
164            Some(RowEncodingContext::Categorical(ctx))
165        },
166        #[cfg(feature = "dtype-struct")]
167        DataType::Struct(fs) => {
168            let mut ctxts = Vec::new();
169
170            for (i, f) in fs.iter().enumerate() {
171                if let Some(ctxt) = get_row_encoding_context(f.dtype()) {
172                    ctxts.reserve(fs.len());
173                    ctxts.extend(std::iter::repeat_n(None, i));
174                    ctxts.push(Some(ctxt));
175                    break;
176                }
177            }
178
179            if ctxts.is_empty() {
180                return None;
181            }
182
183            ctxts.extend(
184                fs[ctxts.len()..]
185                    .iter()
186                    .map(|f| get_row_encoding_context(f.dtype())),
187            );
188
189            Some(RowEncodingContext::Struct(ctxts))
190        },
191    }
192}
193
194pub fn encode_rows_unordered(by: &[Column]) -> PolarsResult<BinaryOffsetChunked> {
195    let rows = _get_rows_encoded_unordered(by)?;
196    Ok(BinaryOffsetChunked::with_chunk(
197        PlSmallStr::EMPTY,
198        rows.into_array(),
199    ))
200}
201
202pub fn _get_rows_encoded_unordered(by: &[Column]) -> PolarsResult<RowsEncoded> {
203    let mut cols = Vec::with_capacity(by.len());
204    let mut opts = Vec::with_capacity(by.len());
205    let mut ctxts = Vec::with_capacity(by.len());
206
207    // Since ZFS exists, we might not actually have any arrays and need to get the length from the
208    // columns.
209    let num_rows = by.first().map_or(0, |c| c.len());
210
211    for by in by {
212        debug_assert_eq!(by.len(), num_rows);
213
214        let by = by.as_materialized_series();
215        let arr = by.to_physical_repr().rechunk().chunks()[0].to_boxed();
216        let opt = RowEncodingOptions::new_unsorted();
217        let ctxt = get_row_encoding_context(by.dtype());
218
219        cols.push(arr);
220        opts.push(opt);
221        ctxts.push(ctxt);
222    }
223    Ok(convert_columns(num_rows, &cols, &opts, &ctxts))
224}
225
226pub fn _get_rows_encoded(
227    by: &[Column],
228    descending: &[bool],
229    nulls_last: &[bool],
230) -> PolarsResult<RowsEncoded> {
231    debug_assert_eq!(by.len(), descending.len());
232    debug_assert_eq!(by.len(), nulls_last.len());
233
234    let mut cols = Vec::with_capacity(by.len());
235    let mut opts = Vec::with_capacity(by.len());
236    let mut ctxts = Vec::with_capacity(by.len());
237
238    // Since ZFS exists, we might not actually have any arrays and need to get the length from the
239    // columns.
240    let num_rows = by.first().map_or(0, |c| c.len());
241
242    for ((by, desc), null_last) in by.iter().zip(descending).zip(nulls_last) {
243        debug_assert_eq!(by.len(), num_rows);
244
245        let by = by.as_materialized_series();
246        let arr = by.to_physical_repr().rechunk().chunks()[0].to_boxed();
247        let opt = RowEncodingOptions::new_sorted(*desc, *null_last);
248        let ctxt = get_row_encoding_context(by.dtype());
249
250        cols.push(arr);
251        opts.push(opt);
252        ctxts.push(ctxt);
253    }
254    Ok(convert_columns(num_rows, &cols, &opts, &ctxts))
255}
256
257pub fn _get_rows_encoded_ca(
258    name: PlSmallStr,
259    by: &[Column],
260    descending: &[bool],
261    nulls_last: &[bool],
262) -> PolarsResult<BinaryOffsetChunked> {
263    _get_rows_encoded(by, descending, nulls_last)
264        .map(|rows| BinaryOffsetChunked::with_chunk(name, rows.into_array()))
265}
266
267pub fn _get_rows_encoded_arr(
268    by: &[Column],
269    descending: &[bool],
270    nulls_last: &[bool],
271) -> PolarsResult<BinaryArray<i64>> {
272    _get_rows_encoded(by, descending, nulls_last).map(|rows| rows.into_array())
273}
274
275pub fn _get_rows_encoded_ca_unordered(
276    name: PlSmallStr,
277    by: &[Column],
278) -> PolarsResult<BinaryOffsetChunked> {
279    _get_rows_encoded_unordered(by)
280        .map(|rows| BinaryOffsetChunked::with_chunk(name, rows.into_array()))
281}
polars_core/chunked_array/ops/row_encode.rs

polars_core/chunked_array/ops/
row_encode.rs
