polars-rows-iter 0.12.0

Library for easy and convenient row iteration of polars dataframes
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132

use crate::*;
use polars::prelude::*;

impl<'a> IterFromColumn<'a> for &'a [u8] {
    type RawInner = &'a [u8];
    fn create_iter(column: &'a Column) -> PolarsResult<Box<dyn Iterator<Item = Option<&'a [u8]>> + 'a>> {
        create_iter(column)
    }

    #[inline]
    fn get_value(polars_value: Option<&'a [u8]>, column_name: &str, _dtype: &DataType) -> PolarsResult<Self>
    where
        Self: Sized,
    {
        polars_value.ok_or_else(|| <&[u8] as IterFromColumn<'a>>::unexpected_null_value_error(column_name))
    }
}

impl<'a> IterFromColumn<'a> for Option<&'a [u8]> {
    type RawInner = &'a [u8];
    fn create_iter(column: &'a Column) -> PolarsResult<Box<dyn Iterator<Item = Option<&'a [u8]>> + 'a>> {
        create_iter(column)
    }

    #[inline]
    fn get_value(polars_value: Option<&'a [u8]>, _column_name: &str, _dtype: &DataType) -> PolarsResult<Self>
    where
        Self: Sized,
    {
        Ok(polars_value)
    }
}

fn create_iter<'a>(column: &'a Column) -> PolarsResult<Box<dyn Iterator<Item = Option<&'a [u8]>> + 'a>> {
    let column_name = column.name().as_str();
    let iter: Box<dyn Iterator<Item = Option<&[u8]>>> = match column.dtype() {
        DataType::Binary => Box::new(column.binary()?.iter()),
        DataType::BinaryOffset => Box::new(column.binary_offset()?.iter()),
        dtype => {
            return Err(
                polars_err!(SchemaMismatch: "Cannot get &[u8] from column '{column_name}' with dtype : {dtype}"),
            )
        }
    };

    Ok(iter)
}

#[cfg(test)]
mod tests {
    const ROW_COUNT: usize = 64;

    use super::*;
    use itertools::{izip, Itertools};
    use rand::{rngs::StdRng, SeedableRng};
    use shared_test_helpers::*;

    #[test]
    fn binary_test() {
        let mut rng = StdRng::seed_from_u64(0);
        let height = ROW_COUNT;
        let dtype = DataType::Binary;

        let col = create_column("col", &dtype, false, height, &mut rng);
        let col_opt = create_column("col_opt", &dtype, true, height, &mut rng);

        let col_values = col.clone();
        let col_values = col_values.binary().unwrap().iter().map(|v| v.unwrap()).collect_vec();

        let col_opt_values = col_opt.clone();
        let col_opt_values = col_opt_values.binary().unwrap().iter().collect_vec();

        let df = DataFrame::new(vec![col, col_opt]).unwrap();

        let col_iter = col_values.iter();
        let col_opt_iter = col_opt_values.iter();

        let expected_rows = izip!(col_iter, col_opt_iter)
            .map(|(&col, &col_opt)| TestRow { col, col_opt })
            .collect_vec();

        #[derive(Debug, FromDataFrameRow, PartialEq)]
        struct TestRow<'a> {
            col: &'a [u8],
            col_opt: Option<&'a [u8]>,
        }

        let rows = df.rows_iter::<TestRow>().unwrap().map(|v| v.unwrap()).collect_vec();

        assert_eq!(rows, expected_rows)
    }

    #[test]
    fn binary_offset_test() {
        let mut rng = StdRng::seed_from_u64(0);
        let height = ROW_COUNT;
        let dtype = DataType::BinaryOffset;

        let col = create_column("col", &dtype, false, height, &mut rng);
        let col_opt = create_column("col_opt", &dtype, true, height, &mut rng);

        let col_values = col.clone();
        let col_values = col_values
            .binary_offset()
            .unwrap()
            .iter()
            .map(|v| v.unwrap())
            .collect_vec();

        let col_opt_values = col_opt.clone();
        let col_opt_values = col_opt_values.binary_offset().unwrap().iter().collect_vec();

        let df = DataFrame::new(vec![col, col_opt]).unwrap();

        let col_iter = col_values.iter();
        let col_opt_iter = col_opt_values.iter();

        let expected_rows = izip!(col_iter, col_opt_iter)
            .map(|(&col, &col_opt)| TestRow { col, col_opt })
            .collect_vec();

        #[derive(Debug, FromDataFrameRow, PartialEq)]
        struct TestRow<'a> {
            col: &'a [u8],
            col_opt: Option<&'a [u8]>,
        }

        let rows = df.rows_iter::<TestRow>().unwrap().map(|v| v.unwrap()).collect_vec();

        assert_eq!(rows, expected_rows)
    }
}